{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# PyCaret 2 Classification Example\n", "This notebook is created using PyCaret 2.0. Last updated : 31-07-2020" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2.0\n" ] } ], "source": [ "# check version\n", "from pycaret.utils import version\n", "version()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 1. Data Repository" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
DatasetData TypesDefault TaskTarget Variable# Instances# AttributesMissing Values
0anomalyMultivariateAnomaly DetectionNone100010N
1franceMultivariateAssociation Rule MiningInvoiceNo, Description85578N
2germanyMultivariateAssociation Rule MiningInvoiceNo, Description94958N
3bankMultivariateClassification (Binary)deposit4521117N
4bloodMultivariateClassification (Binary)Class7485N
5cancerMultivariateClassification (Binary)Class68310N
6creditMultivariateClassification (Binary)default2400024N
7diabetesMultivariateClassification (Binary)Class variable7689N
8electrical_gridMultivariateClassification (Binary)stabf1000014N
9employeeMultivariateClassification (Binary)left1499910N
10heartMultivariateClassification (Binary)DEATH20016N
11heart_diseaseMultivariateClassification (Binary)Disease27014N
12hepatitisMultivariateClassification (Binary)Class15432Y
13incomeMultivariateClassification (Binary)income >50K3256114Y
14juiceMultivariateClassification (Binary)Purchase107015N
15nbaMultivariateClassification (Binary)TARGET_5Yrs134021N
16wineMultivariateClassification (Binary)type649813N
17telescopeMultivariateClassification (Binary)Class1902011N
18glassMultivariateClassification (Multiclass)Type21410N
19irisMultivariateClassification (Multiclass)species1505N
20pokerMultivariateClassification (Multiclass)CLASS10000011N
21questionsMultivariateClassification (Multiclass)Next_Question4994N
22satelliteMultivariateClassification (Multiclass)Class643537N
23asia_gdpMultivariateClusteringNone4011N
24electionsMultivariateClusteringNone319554Y
25facebookMultivariateClusteringNone705012N
26iplMultivariateClusteringNone15325N
27jewelleryMultivariateClusteringNone5054N
28miceMultivariateClusteringNone108082Y
29migrationMultivariateClusteringNone23312N
30perfumeMultivariateClusteringNone2029N
31pokemonMultivariateClusteringNone80013Y
32populationMultivariateClusteringNone25556Y
33public_healthMultivariateClusteringNone22421N
34seedsMultivariateClusteringNone2107N
35wholesaleMultivariateClusteringNone4408N
36tweetsTextNLPtweet85942N
37amazonTextNLP / ClassificationreviewText200002N
38kivaTextNLP / Classificationen68187N
39spxTextNLP / Regressiontext8744N
40wikipediaTextNLP / ClassificationText5003N
41automobileMultivariateRegressionprice20226Y
42bikeMultivariateRegressioncnt1737915N
43bostonMultivariateRegressionmedv50614N
44concreteMultivariateRegressionstrength10309N
45diamondMultivariateRegressionPrice60008N
46energyMultivariateRegressionHeating Load / Cooling Load76810N
47forestMultivariateRegressionarea51713N
48goldMultivariateRegressionGold_T+222558121N
49houseMultivariateRegressionSalePrice146181Y
50insuranceMultivariateRegressioncharges13387N
51parkinsonsMultivariateRegressionPPE587522N
52trafficMultivariateRegressiontraffic_volume482048N
\n", "
" ], "text/plain": [ " Dataset Data Types Default Task \\\n", "0 anomaly Multivariate Anomaly Detection \n", "1 france Multivariate Association Rule Mining \n", "2 germany Multivariate Association Rule Mining \n", "3 bank Multivariate Classification (Binary) \n", "4 blood Multivariate Classification (Binary) \n", "5 cancer Multivariate Classification (Binary) \n", "6 credit Multivariate Classification (Binary) \n", "7 diabetes Multivariate Classification (Binary) \n", "8 electrical_grid Multivariate Classification (Binary) \n", "9 employee Multivariate Classification (Binary) \n", "10 heart Multivariate Classification (Binary) \n", "11 heart_disease Multivariate Classification (Binary) \n", "12 hepatitis Multivariate Classification (Binary) \n", "13 income Multivariate Classification (Binary) \n", "14 juice Multivariate Classification (Binary) \n", "15 nba Multivariate Classification (Binary) \n", "16 wine Multivariate Classification (Binary) \n", "17 telescope Multivariate Classification (Binary) \n", "18 glass Multivariate Classification (Multiclass) \n", "19 iris Multivariate Classification (Multiclass) \n", "20 poker Multivariate Classification (Multiclass) \n", "21 questions Multivariate Classification (Multiclass) \n", "22 satellite Multivariate Classification (Multiclass) \n", "23 asia_gdp Multivariate Clustering \n", "24 elections Multivariate Clustering \n", "25 facebook Multivariate Clustering \n", "26 ipl Multivariate Clustering \n", "27 jewellery Multivariate Clustering \n", "28 mice Multivariate Clustering \n", "29 migration Multivariate Clustering \n", "30 perfume Multivariate Clustering \n", "31 pokemon Multivariate Clustering \n", "32 population Multivariate Clustering \n", "33 public_health Multivariate Clustering \n", "34 seeds Multivariate Clustering \n", "35 wholesale Multivariate Clustering \n", "36 tweets Text NLP \n", "37 amazon Text NLP / Classification \n", "38 kiva Text NLP / Classification \n", "39 spx Text NLP / Regression \n", "40 wikipedia Text NLP / Classification \n", "41 automobile Multivariate Regression \n", "42 bike Multivariate Regression \n", "43 boston Multivariate Regression \n", "44 concrete Multivariate Regression \n", "45 diamond Multivariate Regression \n", "46 energy Multivariate Regression \n", "47 forest Multivariate Regression \n", "48 gold Multivariate Regression \n", "49 house Multivariate Regression \n", "50 insurance Multivariate Regression \n", "51 parkinsons Multivariate Regression \n", "52 traffic Multivariate Regression \n", "\n", " Target Variable # Instances # Attributes Missing Values \n", "0 None 1000 10 N \n", "1 InvoiceNo, Description 8557 8 N \n", "2 InvoiceNo, Description 9495 8 N \n", "3 deposit 45211 17 N \n", "4 Class 748 5 N \n", "5 Class 683 10 N \n", "6 default 24000 24 N \n", "7 Class variable 768 9 N \n", "8 stabf 10000 14 N \n", "9 left 14999 10 N \n", "10 DEATH 200 16 N \n", "11 Disease 270 14 N \n", "12 Class 154 32 Y \n", "13 income >50K 32561 14 Y \n", "14 Purchase 1070 15 N \n", "15 TARGET_5Yrs 1340 21 N \n", "16 type 6498 13 N \n", "17 Class 19020 11 N \n", "18 Type 214 10 N \n", "19 species 150 5 N \n", "20 CLASS 100000 11 N \n", "21 Next_Question 499 4 N \n", "22 Class 6435 37 N \n", "23 None 40 11 N \n", "24 None 3195 54 Y \n", "25 None 7050 12 N \n", "26 None 153 25 N \n", "27 None 505 4 N \n", "28 None 1080 82 Y \n", "29 None 233 12 N \n", "30 None 20 29 N \n", "31 None 800 13 Y \n", "32 None 255 56 Y \n", "33 None 224 21 N \n", "34 None 210 7 N \n", "35 None 440 8 N \n", "36 tweet 8594 2 N \n", "37 reviewText 20000 2 N \n", "38 en 6818 7 N \n", "39 text 874 4 N \n", "40 Text 500 3 N \n", "41 price 202 26 Y \n", "42 cnt 17379 15 N \n", "43 medv 506 14 N \n", "44 strength 1030 9 N \n", "45 Price 6000 8 N \n", "46 Heating Load / Cooling Load 768 10 N \n", "47 area 517 13 N \n", "48 Gold_T+22 2558 121 N \n", "49 SalePrice 1461 81 Y \n", "50 charges 1338 7 N \n", "51 PPE 5875 22 N \n", "52 traffic_volume 48204 8 N " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from pycaret.datasets import get_data\n", "index = get_data('index')" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
IdPurchaseWeekofPurchaseStoreIDPriceCHPriceMMDiscCHDiscMMSpecialCHSpecialMMLoyalCHSalePriceMMSalePriceCHPriceDiffStore7PctDiscMMPctDiscCHListPriceDiffSTORE
01CH23711.751.990.000.0000.5000001.991.750.24No0.0000000.0000000.241
12CH23911.751.990.000.3010.6000001.691.75-0.06No0.1507540.0000000.241
23CH24511.862.090.170.0000.6800002.091.690.40No0.0000000.0913980.231
34MM22711.691.690.000.0000.4000001.691.690.00No0.0000000.0000000.001
45CH22871.691.690.000.0000.9565351.691.690.00Yes0.0000000.0000000.000
\n", "
" ], "text/plain": [ " Id Purchase WeekofPurchase StoreID PriceCH PriceMM DiscCH DiscMM \\\n", "0 1 CH 237 1 1.75 1.99 0.00 0.0 \n", "1 2 CH 239 1 1.75 1.99 0.00 0.3 \n", "2 3 CH 245 1 1.86 2.09 0.17 0.0 \n", "3 4 MM 227 1 1.69 1.69 0.00 0.0 \n", "4 5 CH 228 7 1.69 1.69 0.00 0.0 \n", "\n", " SpecialCH SpecialMM LoyalCH SalePriceMM SalePriceCH PriceDiff Store7 \\\n", "0 0 0 0.500000 1.99 1.75 0.24 No \n", "1 0 1 0.600000 1.69 1.75 -0.06 No \n", "2 0 0 0.680000 2.09 1.69 0.40 No \n", "3 0 0 0.400000 1.69 1.69 0.00 No \n", "4 0 0 0.956535 1.69 1.69 0.00 Yes \n", "\n", " PctDiscMM PctDiscCH ListPriceDiff STORE \n", "0 0.000000 0.000000 0.24 1 \n", "1 0.150754 0.000000 0.24 1 \n", "2 0.000000 0.091398 0.23 1 \n", "3 0.000000 0.000000 0.00 1 \n", "4 0.000000 0.000000 0.00 0 " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "data = get_data('juice')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 2. Initialize Setup" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Setup Succesfully Completed!\n" ] }, { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Description Value
0session_id123
1Target TypeBinary
2Label EncodedCH: 0, MM: 1
3Original Data(1070, 19)
4Missing Values False
5Numeric Features 13
6Categorical Features 5
7Ordinal Features False
8High Cardinality Features False
9High Cardinality Method None
10Sampled Data(1070, 19)
11Transformed Train Set(748, 28)
12Transformed Test Set(322, 28)
13Numeric Imputer mean
14Categorical Imputer constant
15Normalize False
16Normalize Method None
17Transformation False
18Transformation Method None
19PCA False
20PCA Method None
21PCA Components None
22Ignore Low Variance False
23Combine Rare Levels False
24Rare Level Threshold None
25Numeric Binning False
26Remove Outliers False
27Outliers Threshold None
28Remove Multicollinearity False
29Multicollinearity Threshold None
30Clustering False
31Clustering Iteration None
32Polynomial Features False
33Polynomial Degree None
34Trignometry Features False
35Polynomial Threshold None
36Group Features False
37Feature Selection False
38Features Selection Threshold None
39Feature Interaction False
40Feature Ratio False
41Interaction Threshold None
42Fix ImbalanceFalse
43Fix Imbalance MethodSMOTE
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stderr", "output_type": "stream", "text": [ "2020/07/31 01:41:16 WARNING mlflow.tracking.context.git_context: Failed to import Git (the Git executable is probably not on your PATH), so Git SHA is not available. Error: No module named 'repository'\n" ] } ], "source": [ "from pycaret.classification import *\n", "clf1 = setup(data, target = 'Purchase', session_id=123, log_experiment=True, experiment_name='juice1')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 3. Compare Baseline" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Model Accuracy AUC Recall Prec. F1 Kappa MCC TT (Sec)
0Logistic Regression0.82630.89590.72620.81390.76440.62800.63380.0420
1Linear Discriminant Analysis0.82630.89380.75360.79380.77130.63170.63420.0085
2Ridge Classifier0.82360.00000.74990.79200.76800.62620.62920.0134
3Ada Boost Classifier0.80750.86370.70530.78370.73980.58810.59240.0663
4Gradient Boosting Classifier0.80620.88690.73630.76510.74790.59090.59390.1177
5CatBoost Classifier0.80080.88840.72590.76140.73990.57900.58261.7187
6Extreme Gradient Boosting0.79680.88850.72940.75120.73740.57220.57520.0455
7Light Gradient Boosting Machine0.78610.88060.70530.73930.71950.54710.54970.0749
8Quadratic Discriminant Analysis0.76210.82400.62670.73970.66780.48630.50000.0100
9Random Forest Classifier0.76080.83970.66740.71240.68480.49280.49740.1165
10Decision Tree Classifier0.75940.75190.69110.69700.69070.49430.49750.0097
11Extra Trees Classifier0.74330.82050.67080.67580.66980.46050.46380.1478
12K Neighbors Classifier0.72310.76830.60620.66000.62870.40940.41290.0104
13Naive Bayes0.71400.79520.74660.61000.67080.42270.43010.0029
14SVM - Linear Kernel0.52670.00000.42000.24090.25610.02040.02990.0103
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "best_model = compare_models()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 4. Create Model" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.78670.85010.65520.76000.70370.53850.5421
10.85330.93550.72410.87500.79250.68060.6879
20.76000.81930.65520.70370.67860.48750.4883
30.81330.91680.72410.77780.75000.60140.6023
40.81330.88380.86210.71430.78130.62090.6293
50.82670.89660.68970.83330.75470.62250.6292
60.82670.90190.70000.84000.76360.62860.6351
70.84000.93480.70000.87500.77780.65520.6651
80.82430.89120.65520.86360.74510.61490.6286
90.91890.92870.89660.89660.89660.82990.8299
Mean0.82630.89590.72620.81390.76440.62800.6338
SD0.03980.03570.08080.06640.05520.08540.0854
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "lr = create_model('lr')" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.70670.70610.65520.61290.63330.38930.3899
10.74670.72340.58620.70830.64150.44830.4531
20.69330.67990.62070.60000.61020.35750.3577
30.77330.79760.75860.68750.72130.53110.5329
40.78670.78790.79310.69700.74190.56140.5648
50.73330.71850.68970.64520.66670.44490.4455
60.73330.73330.73330.64710.68750.45650.4592
70.85330.84810.83330.80650.81970.69610.6964
80.75680.74140.58620.73910.65380.47020.4777
90.81080.78310.65520.82610.73080.58790.5973
Mean0.75940.75190.69110.69700.69070.49430.4975
SD0.04590.04820.08160.07230.06000.09570.0964
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dt = create_model('dt')" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.79330.86530.60340.81400.69310.54240.5562
10.77330.85340.69490.71930.70690.52220.5224
20.76000.81670.69490.69490.69490.49710.4971
30.84560.89040.79310.80700.80000.67430.6744
40.77850.84590.63790.75510.69160.52070.5252
Mean0.79020.85440.68490.75810.71730.55140.5551
SD0.02970.02410.06440.04690.04170.06310.0625
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "rf = create_model('rf', fold = 5)" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
NameReferenceTurbo
ID
lrLogistic Regressionsklearn.linear_model.LogisticRegressionTrue
knnK Neighbors Classifiersklearn.neighbors.KNeighborsClassifierTrue
nbNaive Bayessklearn.naive_bayes.GaussianNBTrue
dtDecision Tree Classifiersklearn.tree.DecisionTreeClassifierTrue
svmSVM - Linear Kernelsklearn.linear_model.SGDClassifierTrue
rbfsvmSVM - Radial Kernelsklearn.svm.SVCFalse
gpcGaussian Process Classifiersklearn.gaussian_process.GPCFalse
mlpMLP Classifiersklearn.neural_network.MLPClassifierFalse
ridgeRidge Classifiersklearn.linear_model.RidgeClassifierTrue
rfRandom Forest Classifiersklearn.ensemble.RandomForestClassifierTrue
qdaQuadratic Discriminant Analysissklearn.discriminant_analysis.QDATrue
adaAda Boost Classifiersklearn.ensemble.AdaBoostClassifierTrue
gbcGradient Boosting Classifiersklearn.ensemble.GradientBoostingClassifierTrue
ldaLinear Discriminant Analysissklearn.discriminant_analysis.LDATrue
etExtra Trees Classifiersklearn.ensemble.ExtraTreesClassifierTrue
xgboostExtreme Gradient Boostingxgboost.readthedocs.ioTrue
lightgbmLight Gradient Boosting Machinegithub.com/microsoft/LightGBMTrue
catboostCatBoost Classifiercatboost.aiTrue
\n", "
" ], "text/plain": [ " Name \\\n", "ID \n", "lr Logistic Regression \n", "knn K Neighbors Classifier \n", "nb Naive Bayes \n", "dt Decision Tree Classifier \n", "svm SVM - Linear Kernel \n", "rbfsvm SVM - Radial Kernel \n", "gpc Gaussian Process Classifier \n", "mlp MLP Classifier \n", "ridge Ridge Classifier \n", "rf Random Forest Classifier \n", "qda Quadratic Discriminant Analysis \n", "ada Ada Boost Classifier \n", "gbc Gradient Boosting Classifier \n", "lda Linear Discriminant Analysis \n", "et Extra Trees Classifier \n", "xgboost Extreme Gradient Boosting \n", "lightgbm Light Gradient Boosting Machine \n", "catboost CatBoost Classifier \n", "\n", " Reference Turbo \n", "ID \n", "lr sklearn.linear_model.LogisticRegression True \n", "knn sklearn.neighbors.KNeighborsClassifier True \n", "nb sklearn.naive_bayes.GaussianNB True \n", "dt sklearn.tree.DecisionTreeClassifier True \n", "svm sklearn.linear_model.SGDClassifier True \n", "rbfsvm sklearn.svm.SVC False \n", "gpc sklearn.gaussian_process.GPC False \n", "mlp sklearn.neural_network.MLPClassifier False \n", "ridge sklearn.linear_model.RidgeClassifier True \n", "rf sklearn.ensemble.RandomForestClassifier True \n", "qda sklearn.discriminant_analysis.QDA True \n", "ada sklearn.ensemble.AdaBoostClassifier True \n", "gbc sklearn.ensemble.GradientBoostingClassifier True \n", "lda sklearn.discriminant_analysis.LDA True \n", "et sklearn.ensemble.ExtraTreesClassifier True \n", "xgboost xgboost.readthedocs.io True \n", "lightgbm github.com/microsoft/LightGBM True \n", "catboost catboost.ai True " ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "models()" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "['rf', 'ada', 'gbc', 'et', 'xgboost', 'lightgbm', 'catboost']" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "models(type='ensemble').index.tolist()" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Model Accuracy AUC Recall Prec. F1 Kappa MCC TT (Sec)
0Extreme Gradient Boosting0.80220.87900.75010.74640.74750.58500.58570.0349
1Gradient Boosting Classifier0.79550.87690.73650.73910.73700.56980.57070.0912
2Ada Boost Classifier0.79150.86120.70910.74540.72580.55790.55930.0733
3Light Gradient Boosting Machine0.79140.87060.72610.73680.73090.56070.56130.0564
4CatBoost Classifier0.78740.88050.71940.73290.72500.55200.55312.0134
5Random Forest Classifier0.77270.83480.65760.73420.69170.51310.51670.1183
6Extra Trees Classifier0.76600.83030.68490.71000.69540.50580.50780.1597
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "ensembled_models = compare_models(whitelist = models(type='ensemble').index.tolist(), fold = 3)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 5. Tune Hyperparameters" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.76000.83280.79310.65710.71880.51260.5195
10.88000.93330.86210.83330.84750.74860.7489
20.76000.80810.72410.67740.70000.50040.5011
30.82670.91680.86210.73530.79370.64580.6519
40.80000.88230.89660.68420.77610.60120.6192
50.82670.89660.75860.78570.77190.63220.6325
60.80000.90330.80000.72730.76190.59020.5922
70.86670.93410.86670.81250.83870.72530.7264
80.87840.89350.79310.88460.83640.74000.7428
90.86490.92870.89660.78790.83870.72330.7277
Mean0.82630.89290.82530.75850.78840.64200.6462
SD0.04350.04030.05650.07060.04970.08700.0860
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "tuned_lr = tune_model(lr)" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.74670.84630.65520.67860.66670.46250.4627
10.84000.92620.68970.86960.76920.64930.6595
20.77330.79690.65520.73080.69090.51280.5147
30.81330.91600.79310.74190.76670.61140.6123
40.77330.86430.82760.66670.73850.54250.5524
50.78670.85230.68970.74070.71430.54440.5453
60.77330.85330.70000.72410.71190.52510.5253
70.84000.90810.80000.80000.80000.66670.6667
80.77030.85900.62070.75000.67920.50280.5082
90.90540.91840.82760.92310.87270.79780.8008
Mean0.80220.87410.72590.76250.74100.58150.5848
SD0.04530.03940.07420.07640.05980.09520.0957
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "tuned_rf = tune_model(rf)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 6. Ensemble Model" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.74670.82230.65520.67860.66670.46250.4627
10.85330.90400.68970.90910.78430.67630.6912
20.70670.78640.58620.62960.60710.37360.3742
30.78670.86060.75860.70970.73330.55590.5567
40.72000.85310.68970.62500.65570.42070.4222
50.72000.83960.55170.66670.60380.39020.3944
60.81330.86040.76670.76670.76670.61110.6111
70.81330.88110.73330.78570.75860.60670.6077
80.78380.83830.62070.78260.69230.52900.5375
90.86490.89770.75860.88000.81480.70930.7142
Mean0.78090.85430.68100.74340.70830.53350.5372
SD0.05340.03360.07230.09390.07040.11270.1147
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "bagged_dt = ensemble_model(dt)" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.77330.78940.65520.73080.69090.51280.5147
10.80000.84820.68970.76920.72730.57010.5722
20.70670.72340.65520.61290.63330.38930.3899
30.82670.85160.79310.76670.77970.63690.6371
40.80000.85570.79310.71880.75410.58620.5883
50.70670.83620.62070.62070.62070.38160.3816
60.80000.85930.80000.72730.76190.59020.5922
70.81330.87330.73330.78570.75860.60670.6077
80.78380.80270.68970.74070.71430.54070.5416
90.83780.86360.75860.81480.78570.65550.6566
Mean0.78480.83030.71890.72880.72260.54700.5482
SD0.04290.04370.06230.06250.05520.08990.0902
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "boosted_dt = ensemble_model(dt, method = 'Boosting')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 7. Blend Models" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.78670.83660.65520.76000.70370.53850.5421
10.84000.91300.72410.84000.77780.65380.6582
20.72000.78340.65520.63330.64410.41340.4136
30.78670.89810.75860.70970.73330.55590.5567
40.81330.87560.82760.72730.77420.61620.6200
50.70670.84560.62070.62070.62070.38160.3816
60.80000.87220.76670.74190.75410.58560.5859
70.82670.89700.73330.81480.77190.63280.6351
80.75680.86590.65520.70370.67860.48330.4841
90.89190.91190.79310.92000.85190.76750.7727
Mean0.79290.86990.71900.74710.73100.56290.5650
SD0.05270.03790.06580.08710.06640.10980.1113
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "blender = blend_models(estimator_list = [boosted_dt, bagged_dt, tuned_rf], method = 'soft')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 8. Stack Models" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Accuracy AUC Recall Prec. F1 Kappa MCC
00.82670.88530.72410.80770.76360.62740.6298
10.80000.89540.65520.79170.71700.56450.5705
20.78670.83880.68970.74070.71430.54440.5453
30.81330.89320.72410.77780.75000.60140.6023
40.76000.86770.75860.66670.70970.50660.5097
50.80000.87520.72410.75000.73680.57560.5759
60.80000.88560.66670.80000.72730.57140.5774
70.82670.91740.73330.81480.77190.63280.6351
80.81080.87890.65520.82610.73080.58790.5973
90.82430.90270.68970.83330.75470.61980.6265
Mean0.80480.88400.70210.78090.73760.58320.5870
SD0.01960.02020.03410.04770.02050.03740.0377
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "stacker = stack_models(estimator_list = [boosted_dt,bagged_dt,tuned_rf], meta_model=rf)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 9. Analyze Model" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAFlCAYAAAAZGcpRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xd4VGXC9/Hvmd5n0nshCSFAaAFRQGBFighYQFRWXdRVrOu6vK6rD+r66Foeddd1dW2r2NdeEAtWFBEUCL2XEEIqKZMyvZzz/jHJkEhUREJCuD/XlYvMmVPuc64wv7nPuYukKIqCIAiCIAi9gqq7CyAIgiAIwtEjgl0QBEEQehER7IIgCILQi4hgFwRBEIReRAS7IAiCIPQiItgFQRAEoRfRdHcBhBNTv379yM/PR6VSIUkSXq8Xi8XCnXfeyaBBgwDweDw8+uijfPnll+h0OgAmTJjANddcg8FgiO7r3Xff5bXXXsPn8xEMBhk+fDh//vOfsdlsnR77l65/rFVVVXHllVeiVqu58847GTZs2BHt53Cu8a+1ZMkSXnnlFV566aWjsr9LLrmEiooKrFZrh+WLFi06Kvv/MS0tLVx33XW8+OKL0WVLly5l4cKFNDc3EwqF6Nu3L3/5y19ISUnhnXfe4ZNPPuGpp546amWoqanhj3/8I6+99houl4srrriClpYWrr76al555RVee+21o3YsoZdTBKEb5OfnK/X19R2WPfPMM8r555+vKIqiBINB5fzzz1fuvfdexePxKIqiKB6PR7n77ruV3/72t0owGFQURVGeeOIJZc6cOUptba2iKIoSCASUO++8U5kzZ06nx/2l63eHd999V5k7d+6v3s/PXeOj4eOPP1Yuvvjio7a/iy++WPn444+P2v4O1/79+5WhQ4dGX7///vvK1KlTldLSUkVRFEWWZeXJJ59UJk6cqPj9fuXtt99W5s2b12XlWbVqlTJx4sQu27/Qu4kau9AjhEIhqqqqsNvtQKQmKMsyt956a3Qdo9HIggULOOecc/jss88YP348Tz31FO+++y7x8fEAaLVabr75Zj777DMCgUC0pg+ROwA/t/5TTz2F0+nkjjvuAODRRx+Nvr7kkkuw2+2UlJRwwQUX8Pjjj/PNN9+g0+kIh8P85je/4fnnnycxMZF77rmHnTt3EgwGGTVqFDfffDMajYZ//etffPbZZ2i1WmJiYrjvvvtITEyMlvG7777jn//8Jy0tLVxyySW89NJLvP7667z00kuoVCri4+O5/fbb6dOnD7fccguNjY3s37+f3/zmN/z5z3/+Rde4rq6OO+64g/r6empra0lLS+Of//wncXFxTJgwgXPPPZeVK1dSVVXF2WefzY033gjAI488wuLFi3E4HGRlZUX339LSwv/+7/+yfft2JEli7NixzJ8/H41Gw6BBg7jssstYsWIFHo+H66+/niVLlrBz504SExN58sknMZlMP1n+6upq7rzzTioqKlAUhXPOOYcrrriC8vJyLrroInJzc6moqOCll16ivLychx56CK/Xi0ql4vrrr+e0006jtraWv/zlLzidTgDGjx/PjTfeyK233orP5+Pss8/mnXfe4eGHH+buu++Onp8kScybN4+UlBQCgUCHcq1fv54HH3yQQCBAbW0to0eP5t577yUUCnH33Xezdu1atFot6enp3Hfffej1+k6XO51OZsyYwdtvv83//M//UFNTw9lnn80//vEPzjvvPNatWwfAE088waeffoosy6SlpfHXv/6VpKSkDn+fc+bM4ZJLLvnJ6yn0XiLYhW4zd+5cAJxOJ3q9ntNOO4377rsPgHXr1jFixIhDtpEkiVGjRlFcXExGRgYGg4Hs7OwO6xiNRs4666xDti0pKflF63fGZrPx0UcfAfDZZ5/x5ZdfcsYZZ7B8+XLS09PJzc3l1ltvZeDAgdx///2Ew2FuueUWnnvuOaZPn84LL7zAypUr0el0LFy4kI0bNzJx4sTo/k855RRuuOGG6G3elStX8swzz/D6668TGxvLO++8w3XXXceHH34IgM/ni/7emZ+6xh9++CFDhw5l3rx5KIrCvHnzWLRoEZdffjkQ+SL03//+l5qaGiZNmsSsWbPYsWMHn376Ke+99x4Gg4Hrrrsueqy//e1vOBwOFi9eTDAY5JprrmHhwoXMmzePQCBAfHw8b731Fk8//TS33XYbH3/8MQkJCZx33nl88cUXzJgxA4AHHniAJ554Irrf+fPnM378eG666SZOP/10LrvsMlpaWrjoootISUlhyJAhVFdX8/e//50RI0bQ1NTErbfeyrPPPkt6ejo1NTWcf/759OvXj3fffZf09HQWLlyIx+NhwYIFtLS0cN999zFjxgwWLVqE0+mkoqKCoqKiDtdSkqRO/05efPFFbrjhBk4++WTcbjenn346mzdvxufzsWrVKj766CMkSeLBBx9kx44dyLLc6fK2L3g5OTn87W9/4+6772bRokWUl5dHj/Xee++xc+dO3nzzTTQaDa+//jq33XYb//nPfw75+xROXCLYhW7zwgsvEBsby5YtW5g3bx4nn3wycXFx0fdDoVCn2wUCAdRqNSqVClmWD/t4v3T9zrT/snHeeefx7rvvcsYZZ/DOO+9w/vnnA/DVV1+xadMm3nrrLSASvgBJSUkUFBRw7rnnMm7cOMaNG8eoUaN+8njffPMNZ555JrGxsQDMnDmTe+65J/phP3z48J/c/qeu8dy5c1mzZg3PPfccpaWl7Nq1iyFDhkS3Pf3006PljouLo6mpiZUrVzJp0iQsFgsAs2bNij5fX7ZsGa+++iqSJKHT6bjwwgt54YUXmDdvHgBTpkwBIDMzk/z8fJKSkgBIT0+nqakpetybb76ZM844o8N5eDwe1q5dy8KFCwGwWq3MnDmTZcuWMWTIEDQaDUOHDgUiNeja2toOXzokSWLHjh2MHTuWefPmUVVVxejRo/l//+//YbVaOxxfpYq0KT7cv5X777+fZcuW8eSTT1JSUoLf78fj8VBQUIBarWb27NmceuqpTJkyhcGDB9Pc3Nzp8vYB/mOWLl3Kpk2bmDVrVrSMXq83+n5nX4aFE49oFS90u4EDB3Lrrbdyyy23RD/cioqKWLNmzSEfrrIss3r1aoYNG0ZeXh6hUIjS0tIO6/j9fq688kpqamo6LD+c9SVJQmk3fUIwGOywbvvbxVOnTmXDhg3s2bOH1atXR8NIlmUeeeQRFi1axKJFi3jzzTe54447UKlUvPzyy9x33304HA7uvfdeHnjggZ+8Np2Fi6Io0S89P3f7uk1n1/jBBx/kkUceISYmhgsuuIAxY8Z0OHe9Xh/9vf11ab+OWq3uUFZJkjq8bv/lTKvVdvr74ZBlucNxf7h/nU6HRhOpp4TDYXJzc6PXf9GiRbz++uuceuqpDB48mC+++IILLriAiooKZs+ezebNmzvs1263k52dzYYNGw4pxx//+Ee2b9/eYdnFF1/M119/TU5ODtdddx2JiYkoioLNZmPRokX85S9/Qa1Wc+ONN/LKK6/86PLDvQ5XXHFF9LzefvttXn311ej7h/v3IPRuItiFHmH69OkMHjw4ept4ypQpGI1G7r333miN1+fzcffdd2M2m5k0aRI6nY4rr7ySBQsWUFdXB0Rq8/feey9erzdaI2xzOOvHxMSwZcsWFEXB5XKxdOnSHy2zXq9n2rRp3HLLLUyePBmj0QjAqaeeyvPPP4+iKAQCAa655hpefvlltm/fzvTp08nNzeWqq67i0ksvZdOmTT95XcaOHctHH31EQ0MDAG+//fYhz7aP9BovX76cuXPncs455xAXF8eKFSsIh8M/uY9x48axZMkSmpubkWW5Q2v1U089lZdffjl63m+88QajR4/+xeXsjMViYciQIdEAbGlp4b333ut0/0OHDmXfvn2sXr0agG3btjFlyhRqamp46KGHePzxx5k4cSILFiwgLy+PXbt2odFoCIfD0S8P119/Pffccw/79u0DIl8WHn/8cbZv305OTk70WM3NzWzatImbbrqJyZMnU11dTVlZGbIss3TpUi699FKGDRvGH/7wB8455xw2b978o8sPx6mnnspbb72Fy+UCIu0dbr755iO/sEKvJG7FCz3G7bffzllnncU333zD2LFjWbhwIY8//jgzZ85EpVIRDoeZMGECCxcujNb4rr76aoxGI7///e+BSO175MiRPP74450e4+fWbzv+5MmTSUpKYuTIkYfUFNubPXs2L7/8MnfeeWd02YIFC7jnnnuYMWMGwWCQ0aNHc8UVV6DVapk6dSqzZs3CZDJhMBi47bbbfvKajBkzhksvvZS5c+ciyzKxsbE89dRT0dvFv1T7a3zdddfxwAMP8Mgjj6DVaikqKqKsrOwntx8/fjw7duxg1qxZ2Gw2CgoKog3RbrvtNv72t79Fz3vs2LFcffXVR1TOzjz00EPcddddvPPOOwQCAWbMmMHMmTOpqKjosF5sbCz/+te/eOCBB/D7/SiKwgMPPEB6ejpz587llltuYfr06eh0Ovr168e0adNQq9UMHjyYadOm8corrzBjxgwURWH+/PmEQiH8fj8DBw7khRde6NAg02azMW/ePM4991xMJhNJSUkUFRWxb98+Zs+ezbJly5g+fTomkwm73c7dd99NSkpKp8sPx+zZs6NtBiRJIiUlhfvvv/+oXWOhd5CUn/rUEgRBEAThuCJuxQuCIAhCLyKCXRAEQRB6kS4N9g0bNnQ6SMKXX37JrFmzuOCCC3jjjTe6sgiCIAiCcELpssZz//nPf3j//fejLYXbBINB7rvvPt566y2MRiNz5szhtNNOIyEhoauKIgiCIAgnjC4L9szMTB599NFDumLs2bOHzMzM6LCWw4cPZ82aNUydOvVH9yXLMm63G61W26GfrCAIgiD0RoqiEAwGMZvNv7gXTJcF+5QpUzodScnlcnWYuclsNkf7ZP4Yt9vNzp07j3oZBUEQBKEny8/PP2S2w59zzPuxWywW3G539LXb7f7ZQrf1Wc7Pz+/Qh1Q4ujZv3kxhYWF3F6PXE9e564lr3PWOp2sclmWafSGa/QGavEGafG3/Bjssa/YFCYZaCIU9hBUvGsmHSevDrg/hCWpYvCMBWZGYObCKk9JaUHVyA7nGpcOgUbDqg52+H6FBo7Kj01pREcCgNRNnTkTlN3LpnKtJSEjgrbfeoqSk5BeP0hjZ+zGWm5vLvn37aGxsxGQysWbNmuhgIT+m7fa7TqfrMMylcPSJ63tsiOvc9cQ17nrH8hqHwjKN3gDOth9PAKfXj9MboNHTflmARq8fp8ePK+Cj1hWiyRdCp5Y5Ka0JuyFIvClIjDGITR/Goguxq9bGqxtTAHjojB3EGDufpyI/ScGgDaBVNRImTLiTUWBizR5MOhtmfTxatQGbMQ67MRGrMRaLPgaL3oFG3bGCGg6Ho8Mzz58/n0GDBkWHBz6Sx8/HLNgXL16Mx+Phggsu4JZbbuH3v/89iqIwa9asQ4b+FARBEHqfYPtw9vijYdz4U2HdurzFH0SnlrHqQ1h1YbRqmV31ZgCyHV4m960l3hAixxbCopMxasNoVPB1aR/2NfchwQxTc7d1Wq5Jec1cPDwHRXbR4NYSDHce7HZ9FSpJjUlnw6S3YzPEYTHEYNY7IqFtcGDS2VGrDi9aA4EAd9xxB+Xl5bz00ktIksSll14KREbFPFJdGuzp6enR7mxtUzICTJgwgQkTJnTloQVBEIQuEAzLOD1+9jX7Ce6rPVhLbhfKB2vOHcPa5e8YmFpVa1Drw1j1IWz6MFZdiANuHeuqbFj1Ws4uqGVk+gEMmhBq1cEqsqKoCEiTsJmyMWoqaGz+oNPynjvQT58EHS1eJ2UNVgIhD4oio9B+XyGqG9cAoFZpsRsTsRgcmFtr2JHQjgS4UWdFJR2dnuIajYbt27dTVVVFfX098fHxR2e/R2UvgiAIwnEjEAp3COC2GnTHmnPny9yB9uG8p9P9a1QyBfFurPow2XaFEakKMYYwNkOYClc2ISWdGJOOvvYlaFUth2yvVmk5a+hvsZtiWLP3Izb/YD4AAEmSGZLSQrylgSavCyWYQCDkIyyHCIZ90eBucFfS4K6MbqdTGzAbDg3sthq3XmPu0t5XHo+H4uJixo4di0ql4umnn8ZqtR7SNfzXEMEuCIJwnFAUBV8oTIsvSIs/RIs/iKv137Yflz9Esy8YqSUfUnOO1J49gZ+exa+NRiVj1YWJN6vQqG3kJ9jIcgTpn3AAreIi1qJGrwmgUflR4SPefi5J9iyseoXvdv29033mxLj57ajZAHy5dStlDVs6Oa6WZm81Ln8tVkMcCdZMZDlMUA4QCHnwBSMNsLdXrThkW4PWQpwx/Qc1bgcWQyxmvQOdxnC4l/uoUxSF2bNns3btWr7++mvy8/NJTEw86scRwS4IgtBFOgviH4axq7OA9kVeuwKt2/la1w2ECMu/fN4uSQK7QUecScPwNAOJFog3ycQYZfRaC3ptJg6TDotmN1J4J+AjLHsJy5HnvFq1gXOK/oRZb6fCuZPPtnzX6XF00jpOyhyOoijsrcmhprkk+p5GpcOos2LRO/AHvXgCTWTG9kej1hJqrWX7Ax7cgSb8ITdfbHvh0PNAwqS3kWjLbg3rGCz6gzVus96BRv3LW5EfK5Ikce2117Jy5UoyMjK67Dgi2AVBEFr1lCBuo9eosOq1WPVasmIsWPQaLHotVr0mstygxaJTEWPwYNYFMWqD6NUBtKpILTo9tpCc+H7YDTo+2fw0Nc17DzlGrDmFKYW/Qa81UVy6l03l1UiooN0z6GDYx97aDRSmjyPemkGsOZUGdyUqSYVRa8Woi/wk2bJRFAV/yM2A1DGkxuQRDgfxhzx4As24/I3Uuyp49fv/7fR8VZIas95BrDmltcbd8Xa5WWdHpVIf8fXsDqtXr+bhhx/m2WefxWg0Mm3aNKZNm9alxxTBLghCrxEMy9S5fex0+mjYUUmt20+dy0dzjwpiTXQdiyHyu0WnwWrQYtapMGrCxJntWPRa/MFmyhq24Au68AXd+AIN+EJufEE3p/efi92UQEgO8vKK2wFQguALgq+1PAa1l+GZg1vLZkQlqZGVjrfhG9xVNHlrSdRmMSRjAtsql6NW6TDqLJHA1lox6azEW9MB0Kn1jM6byaat68nMTsMdaMTlb8Ttd7Krppj1ZZ8TkoOdXh+NSofF4CBBnxV9pt2+gZpRZ0E6Sg3Teor33nuPTz75hC+//LLLA72NCHZBEHosWVZo8Pg54PJR6279t8XHAZevdZmPWpePA63LnN5Au61LfnS/bX5tELctt7Zup1V3DCVZCeMLevAH3fiCLtQqHYm2TAD21m5kb90qfEE3LS43tU4XgZAXrdrARaPuBKCupZ5VJYsPKbdK0rC5/GvG5J+HRqUlwZpJbUvZIetVNe1GVsKoJDXDsibj9NREw7p9cFsMMQBo1FouOPk2fEEXLl8jLr8Tt78Rl8/J+rLPcfsbcfubol8O9u364fU0YTMmYNE7Om2gpteYTohhwUtKSsjJyQFgwYIFzJgxg1NOOeWYHV8EuyAIx4yiKDT7gtS6D4bxAVcknGvdfg60tAZ1NLT9yMpP16AlCeLNelLtRoamxRBvNoC3mcKcTBIsBuLNemytIf1zQXw45Y/WnoPN+IIu3F439c0uUhx5JNsjH+Zfbf8vVY278Yc8HbZPtucwceBlrY3D6iirjzQcU0saJEmFVq1HUcJsLl9GYfo4YszJxJhTcLqrOuxHVkI4vTXR14Xp49lzYC0GrQWj1oxBa8aos2HUWoFIkMaYk5k14s+EwgFc/tbQ9jXS4qtnzd6PcPmduHxOPIEW2t+Gb8+otRJrScWij8HV6Ccvq6BDjVurEYMCPf/889x00008//zzTJ8+HZPJdExDHUSwC4LwKyiKwpr99ayvdHZY1uQNRkP7gMtHXbsgD4Tln92vw6gjwawnL85KgtVAoiXyk2A2kGAxkGhte60nzqxH/YNJMoqLixk+fPBhlF+O3vr1BV1UNZa0hnZbeEd+H5V3Lg5TIgoyr6+6p9N9NXnrosEeDgcJy2G0agMKCrIcRlZCVDeVUNeyn2R7DgWpo9hU/hUhOUBYCYESecZs1Fmi+zRozQxKH4/H34RBa4n86MwYtRb0WnN0vay4gWTFDQTAH/Li9jlx+RtpcFdQ1rAFV+trt98ZbVH+QxIqTHobSa0N06I17tYGaia9HY3qYMO04uJiClKG/+w1PtGccsop5OXlERcX121lEMEuCMIvtrW6kdfWlfL6+lJ21x3aD/mHTDo1iRYDQ9NiIsHcGtKJVsOhr816dJojayClKDL+kBef3Eyj5wAOU6QrUWXjLsrqt7YL7Mi/ITnIxaPuAqDJU8fXO/7b6X53Vq9iZM50VJKaBGsmDe4qwnKI9jXb0rqNhOXzUas0FGWfwaJ1DwMSBq0Jg96BQWvGoLWgVUdqtXqNkdP6X4xWre/w3g9vVeckDG13fgq+oBuX30ltS9kht8tdfifBcOcjlqkkDRa9gxhzSnRo0/Z9uE16Gyrp+GqY1hP4/X7+8Y9/cNFFF5GZmUlBQQErVqz4xTOyHU0i2AVBOCylDS5eX1fKa+tK2VgVqaGbdGouGJrNpH4p6Nrd1rYbde1q2XrM+iPrghSWQx2e6wbDfnxBN3GWNBKske5C35cspqpxN76gG3/QHR2YxLVnN1MGXQlAg6uS7VUro/tVSRrUKg0qSc2Oqu/pl3IydlM8DlMSjZ6aQ8pR03SwNXlByimsL/siGsbG1lq0QWtBaX1sYDfGc8HI29BrTT85SllaTH6H17Ii4/E3dRrYkRp3I+EfaZimVes7DLRi7hDeMRi15l7XMK0n+OSTT3jwwQfZv38/jz/+OEC3hjqIYBcE4SfUtHh5c/0+XltXysp9tQBo1SqmD0jnwmHZnDUw/YhDGyI1UG/QhcvXQIuvAZevgdzEIiyGGGQlzCsr/3pIK26AQemnRYO9xVtPi7cBjVqLQRu5jR0IBqhzVVDVuIcURy45icPYuH8pgXCkvbishJBbxwNvC3KD1kLfpBE0uCox6Fpve2sjt71NOlv02LmJReQmFv3kealUHW+pt4l8UWlqfb7tPPisuzXA3YEmFKXzRxV6jQnHD4Y6bd9ATacxnhAN03qCQCCAJElotVpmzJjBgw8+yPnnn9/dxYoSwS4IQgeN3gDvbCzjtXV7Wbq7BllRkCSYkJfMBcOymTk4k1jT4TeSCob9uHxOWnwNZMT2R5IkGj0H+Gr7y7T4nIfUQG2mRCyGGFSSmmR7Lv6QG5WkIhQOEpIDBMMBtlZ+w+CM09CqdYzocybvrf0H4dDB/UioMGrsrbfLwaSzMSxrCpIkRWrY7WrZOvXBkcgGpo39VdcuGA7g9jujNe62Bmpt4f3jDdOkSJcyS3pkoBVDa807Wvt2RG/hC91r7969/O53v2PGjBncfPPNSJL0szOUHmsi2AVBwBMIsXhLOa+t28uS7ZXRBm4nZ8Zz4bBsZg/NIsVm6nRbWQnjD3ow6qwA1DSXsr1yZaQG7m/o0Fhr1vC/YDVGapctPic6tQGV1oyEFL3VvqHsM/rEDwIgJ2EIy3e92e5oEgatGYcxiWDYh1atw2KIYXTerNYZt2yYdDY2b9jGiBEjOpSzf+qoX3WNFEUhEPYebIjWrkGaq7XG/cNW8NFSSyrMOnukYVprY7T2NW6z3nHYM4IJ3SsuLo7Gxkbq6upQFKVH3iURf0mCcIIKhMJ8urOK19bu5f0t5dHJPQqTHVw4LJsLhmWTE2ftsE1Ncyk1TXtbb5s7W8PbiUln5axhN2DQWvAF3eyt2wBIaFTa1i5cMiE5xPr9nzE2/3yMWgt5icPZUX1waFKtWo9JZ8duPDh2doojl9MKLo4GtlFrPWTkMY1KS37ySR2WHcmHbdtjgbYat7u1xt3+efePNUxTqzSY9THEWdI6rXEbdbajNiOYcOxt3rwZt9vNySefjM1mY/ny5djt9u4u1o8SwS4IJ5CwLLOs5ACvrdvLOxvLaPBEBnTJibNwcVEcp+cZiTMFaPGVsrdmHZv2NZBgy2R03kwA9tdvZXPFsuj+Io2xFDyBZsrqt5GffBKpjjwcpkQaPQcIyYF281fboqEtSRIFKaeQHT8oGtqd3Wo2t9ZmjwZZCePxN/+gUVpjh99lpfN5uLVqfaczgbU96zZoLT2y5ib8enV1dUyZMoX4+HhWrVqFXq/v0aEOItgFoddTFIVVZbW8s3Eb35aUoKKFeHOQiwaHcZjzmTpwDCMz4/l401Psri5ld7tt1SotFr2DYMiPVqMnK66QzRXf0PacWEIiwZZNnDk12rVMq9Zzat/zUakigf5jo43FmJOP6nm2taB3+Z00hPaydl99hwZqHn8zCp03TDNozcSYk6MzgZl/cLtcrzl6U2oKx4e22+zx8fEsWLCA/Px89Prjo52DCHZB6CX8IU+HW+RBJZt3Nzfw+rpS5o9aTf8Ymf4/GE9kYJqWk7ISAOibNIJURz4QGV2txVtPnauc/c7tVDbuIiu+kARbJv1TRkVn2IqzpHUYtKRN27jiR1Mw5G/X7evQBmreYMf+9BX7236LNExLsGV0nAnMcHBGMK1ad9TLKxyfQqEQjzzyCFu3buWZZ56Jzsh2PBHBLgjHibAcioS2v4E0R18kSUWzt56vtr8SCfKwr8P6j3+fQXGlDZNOTVMgEZvJSt+EVOymeKyGGCyGWLRqPWE5hFqlITNuAG+sujfakhzAoo8hLSYfQ7uuWyfnnnXUzy0yI5jn0H7b7W6X/1jDtMiMYHaSTTnRoK6vbmZgv6FYDA5MOrtomCYcNpVKxVdffUVJSQk1NTUkJx/dO0vHgvhrF4QeqsXXwL66zdS59lPXUo7L30jbLfDzT/ofTHobOo2BRs8BgrKF8mYTu2qh1qPD6dWTHd+H+RP6MWNApK+5oii4/Y0caCljX/0WappLcbqrmVx4OamOvug1JlIdfTHrHSTZskm0ZWPWH51niYoi4w24Dqlxt2+gFpIDnW7b9jgg3preYVzythq3UWc9pGFacV0xKY7co1J2ofcLhUJs3LiRoqIqcmltAAAgAElEQVQiVCoVTz31FCaTCYfj6LTvONZEsAtCN1MUmSZvLbUt+3H5GhiWNRmARnc1a0o/AkCvMZNs74NFH4PVEEuzP8Sr63fx+rq9LN3dD1kBlSRxWl4SFwzPZuagTBzGyO1lSZLwBlp4p/ihDq261SoNSbZs2iYJATh9wNwjOgdZCeP2N3WocUd+P9iivLOBZgB0agNWY9zBwG43PrnF4ECvMYuGaUKXuvjii1m2bBnLli0jLy+P1NTU7i7SryKCXRC6gdNdzd7aDdS27KfOtb9D4A5MG4dOYyDRns24fnNIsEaeDXsCIRZvLee15aV8sv3jaF/zU7Ja+5oPySbeoqG6sYTtlR9T7tzGaf0vId6SjkFrwW5MwGKIId6SHn0+fri3qEPhYLRhWtu/B0O7rWFa5zOCGbQWYs2pP5h7+2ADNZ3G0Ol2gnCszJkzB4fD0a0TtxxNItgFoQuFwgHqXZXUtpThD3kYnn0GAA3uKjaWLwXAZkwg05pBvCWDBGsGGrWWsCyzzxliU5WDTcX72VC5gS92VUf7mg9Kae1rPjSbjBgDe2s3sLXiLSqdu6K3tHUaIy6fk3hLOpIkMX3o9T9azkDI1y6wD22g5gu6Ot1OQsKkt5Fgyzpk7u222rdGfeRDzgpCV9i5cyf/+Mc/+Oc//4nBYODss8/m7LPP7u5iHTUi2AXhKKtp2sue2vXUtZThdNdEu1ipJDVDMycCkBbTl0kDLyfemk6TV2JTVSOf7HKyqaqMzdUb2VLdiDfY8dZ1TpwlGuZpNj9GnRWD1kxYDrGqZDHBsB+bMZ6M2P5kxPYn0ZaFSlLj8jVS59qPLIdRWlu8u9qPU+5zRsdQ/6FIwzQHMabk6NCm7W+Xm3X2QwaMEYSe7vnnn+eNN95g4sSJnHfeed1dnKNOBLsgHCFvoCVyK71lPwpKtDbu9FSzs/p71CoNCdYM4q3pxFszsRpSWFfeyAd7Gvlv+XY2VTnZVLWBA66OoapTqxiQZKcwJYZBKQ4GpcQwMMkCShXlzu1sL/+UNX4np+SeTUHKKNQqDWPzL8BuSsBuTMATaKa6sYSVu9+luqmEFl/Dj56DRqXDYnCQoM/qMBNYW+3bqLOIGcGEXqGmpoakpCQAFixYwPjx45kyZUo3l6priGAXhF+grH4rJbXrqG3Zj9vfGF2u0xgpap1kJCNmIL7sWEoadBRXt7BxvZPNVaXsrtuErLQ9h64EIDvWzPQB6QxOdVCYHAnyvHgrGrWESlKhKAqfbn6WL7eWRkdF06oNZMcPjo7i5gu6kJUQWyuWU9VUQrO3NlourdpARmx/kmx90Goig2sYNOZoi/IfGzxGEHqTt99+mz/84Q8899xzTJkyBbPZ3GtDHUSwC8KPqndVsLN6FVqNgRHZU4HIFJ+ldZswaM2kxxRg0qfQ6Hewp0HPvDe+Y3O1ky3VTdFn4W0cRh1j+iQwKCUGW6CZaScPpjDFgc2gIxDyUecqp65lL1UNZWzct58R2VPJSxqOJEUmR3GYEkm29yEjtj8OYxIHWvZRVr+F70ve7zB/uEatIy2mHyn2HJLtucRaUsUY5cIJr3///sTGxnZ3MY4ZEeyC8APBcIDtVSsoLl0CgM0QT2HaZLbVNLGxOp6djWeytsLP5uomqpoPAAei22rVKgoSbQxqvY1emBLD4BQHaXZTa0jLrC0uZkSfRGQ5zKK1/8TpqaH9VJ4mnb1D17AJ/S/hQMs+qhv3sHrvRzS4q6Lrq1VaUhx5pNhzSbbnEm9JE8+8hROeLMs888wznHnmmaSnpzNgwADWrl2LTndijDAogl0Q2qltrmDp9v/iCdQDoChw59I8Lnz9VcJyx+5cGQ4TZ/ZP6xDg+Qk2dJqDwerxN1Pbspfi0v3UtpRR76ogVROZTlSlUqNR60iyZZNgyyTBkkGCNROtxsCB5lLWlH5MdWMJ9a6KDg3wkmzZpDhySbbnkGDNFKOqCcIPLFmyhFtuuYVVq1bxzDPPAJwwoQ4i2IUTmDcYYnVZfWsjtkY2VzUSZ9zJuf3r+XR3LCvLHNR7tKjVPk7JjO/QmK0wxREdAKZNKBxA1fq4OhDy8d7ah/EEmtqtIUUmSgkd/IJw5uBrCMshalv2UdVUwpbKb6hrKY/W2CVJRYI1g2RHDin2XBKsWaL7mCB0QpZlZFlGo9EwdepUbr/9di6++OLuLla3EMEunJDWltcz+4WvKW1wY9MH8QTVKKgpSMzi+8psBqX34bcjYxiU7CAz5tCRzxRFptFzgNqWMupaIrVxp7uGyYWXk+LIQ6cxYNJZibOkkmDNJMGaQZwlHZ3GwOo1q6lp2ktV0x6qm0o40FwWbRgnIRFnSSPZkUuKPZdEW1an05kKgnBQRUUFV199NePHj+emm25CkiT+9Kc/dXexuo0IduGE88LqPVz71vf4QmH+PN5KQex6Eu2DOL3/Oeg1nT+f9gXd6DQGVJIat7+J99Y+3GHSlbaube1v1rcNCCMrYepaKthetZLqphKqfCVs2dT2DF0i1pwSaezmyCXJ1keMxCYIv5DZbKakpISEhITodKsnMhHswgkjEArz/94v5vFvdxBvUvPseSp8/u9QSRr6xCVHQz0sh2hwV0b7qNe27KfFV8/0odcTb0nHpLNiM8bjMCYQ31objzEnR591y4pMg7uS6sYSqpv2UN28l1D44AQneslGTnIhyfYcku056LWmbrkegnA827dvHw0NDQwbNgyHw8EXX3xBUlLSCR/qIIJdOEFUNnm44MVlrCitZXyOlqtPKsfjr8NuTGBEn2mkx/QDwOmuYfH6f3Vola7TGCPzlLf2QZckFTPaDc+qKDJOTw3VjXuoatpDTdPeDiO52YwJ0e5nyfYctm7awfDcH0yMLgjCYWtsbGT8+PHExMTw7bffYjKZjsvpVbuKCHah11tecoALXlxGdYuX3w1P5rSsr/EEQtiNCfiCbr7Y+jwzh9+EzRiPzRhHnCWNOEta66hxGdgM8R1qAYqi0OStpapxT6RG3lTSYa5wqyGWrPhCku2R5+Qmva07TlsQei2Hw8ENN9xAWloaRqOxu4vT44hgF3otRVF4/NsdzF+0BgX4+1nDufKUNJZu20G9q5Imby0mnY3s+MHISqQ7mVqlYdqQaw/ZT7O3LtLYrbGE6qYSvMGW6PtmvZ302CKS7ZGW6xZDzLE8TUHo9RRF4YUXXmDNmjU89thjAMyfP7+bS9VziWAXeiVPIMQ1b33Py8UljM70c8MYLecMzeGtNQ/gD3mwGmIZmjmJPglDOh2ZzeVztgb5HqqaSjp0WzNqrfRJGBIdFMZqiBXP9QShCymKwhtvvMHWrVu55ZZbSE9P7+4i9Wgi2IVeZ299C+c9/zWbq+u5cUwjgxKr8PrVNPvqKUwfh05jpG/iiA4jtLn9TVQ3RRq7VTWW4PIfnDhFrzGTFTeIFEfkObndmCCCXBC6mKIo7Nixg4KCAlQqFU8++SQajYbU1NTuLlqPJ4Jd6FU+2V7JRS9/g17dzL+mVaLXeFFJas4cfE302TlEZmarbiqhqjXMm7110X3o1AYyYgeQ0tqX3GFKFDOcCcIxdvXVV/Phhx+ybNkycnJyyMzM7O4iHTdEsAu9giwr/N+Xm7l9yXrO6lfLjP4HkAAJFYVp4zDqbOyr29w6KMweGj3tx3fXkx5TEHlG7sglxpwiJk4RhG42efJk6uvr0evFAE2/lAh24bjX7Atw6asrWLKtlMuLnIzOjExbGmNKJsGWxX7ndjaWf0XbxCkalZZUR99Iq3VHDnGWNFSSmDhFELpTZWUljzzyCHfddRd6vZ6ZM2cyc+ZM8djrCIhgF45boXCA1fs289Kqb4jVeZmQN5AJuZVIWDBozTg9NTg91agkDcn2PgdnQLOmi4lTBKGH+fe//81//vMfCgsLueSSS0Sg/wri0004roTkIFsrllPZuIvqplJAZmRrA9l4y348AQlPoBlvsIUEayYFKaPIiisUE6cIQg/U3NyMzRYZ5+HWW2+lsLCQCy+8sJtLdfwTDxKFHq3FV8+Oqu+jjdvUkpqtlcupbiqhyScRmUlVQoWaOlcZvqCbvMThTB9yPdOGXEtu4jAR6oLQA3366acMHTqUzz77DACLxcKcOXNETf0oEDV2oUfxBd1UN+2hsnE3lc7d0W5nI7KnUpg+njqXn+/2J5EXsw+7oW3YVwWDzkL/1FH0TToJg9bcfScgCMJhaeu25nQ6u7kkvY8IdqFbheQgKAoatQ5ZCfP2mgcIhv1ApNtZVtxAUhx9SY8pYOXe3azY9RJDk/3R7ZPsOQxMHUN6bH/Rkl0QejBFUXj77bcZPXo0qampFBYWsnHjRiwWS3cXrdcRwS4cU4oi0+CupqpxF5WNu6lp3svJOWeRnzwSlaSmIGUUGrWOVEdf4ixpSEjUufbz1tq3CAZ3E2cCRZHokzCEoZkTcJgSu/uUBEE4DJ9//jnz5s1jxowZvPDCCwAi1LuICHbhmFAUhW92vkFl4058QXd0ucOU1KGF+vDsM4BITb7kwDrWl32Jy18PQINXT07iCM4aNBG9Vkz8IAg9naIoKIqCSqVi4sSJ/PGPf2Tu3LndXaxeTwS7cNSFwkEqGndS1biLzNiBpMb0RZIkWnz1qCQ1uYlFpDr6kuLIw6SzdtjW5XOyo/p7dlR9F536VFbg41353DtjFn0T7N1xSoIg/EL19fXMnz+foqIi/vjHPyJJEn/961+7u1gnBBHswlFV3VTC8p1vtRtrXSI1pi8AEwdcik5jPKTVq6IoVDftYVvlCvY3bENpHUgGYK/TQLXnJJ6/aDJmvWjdLgjHC5VKxapVq2hubuYPf/gDKpVoA3OsiGAXjopgOMDa0iVsq1qBhET/lNFkJwwmwZIRXUevNXXcJuRnT+1atlWupMkbGeLVqLXiDbYQDEu8timFswZP5M7p/UUXGEE4DtTV1XHgwAEGDBhATEwMH330EVlZWSLUjzER7MJRsa3yW7ZVrcBuTODU/NkkWH98woYmTy3bq1ay+0AxwbAflaQmJ2EoBSmj2VhezMaKdTz2XToPnj2FWYOzjuFZCIJwpJqbmxk7dixms5lly5ZhMpno06dPdxfrhNRlwS7LMnfeeSc7duxAp9Pxt7/9jaysgx/Szz77LB9++CGSJHH11VczadKkriqK0EVC4QBqlQZJUjEg7VQkSUX/1NFoVIfeMpcVmfKG7WyvWkll4y4ATDobhWnjSLbnkmjLwukNcONiiX3OHO6ddrIIdUE4jthsNi6++GIcDoeYuKWbdVmwf/755wQCAV5//XXWr1/P/fffzxNPPAFEvtm99NJLfPrpp3i9Xs455xwR7MeZmuZSvt35Fv1STmFg2qloVFoGpY8/ZD1f0M2umjVsr1qJ298IQJKtD/1TR5MR05/dtcV8svk/DEqfyJ8+CLDtQAvzxw/iD2P7H+tTEgThF/rggw9Yvnw5999/PwALFizo5hIJ0IXBXlxczNixYwEYOnQomzdvjr5nNBpJTU3F6/Xi9XrF89PjSCgcZN2+T9hS+S0A/pC70/Wc7mq2Vi6npHY9YTmERqUlP3kkBSmjiDWnEAoHWbnnXXYfKEanMfLMqga+KfEwe0gW/ze96FiekiAIR0CWZR599FE2btzIVVddJW679yBdFuwul6vD4ANqtZpQKIRGEzlkSkoK06ZNIxwOc9VVV3VVMYSj6EDzPpbvepNmbx1WQxyn5s8myZZ9yHqyEubDDY8TkgNYDXEUpJxCXtJw9JpI47kWXz1Lt71Mg7uKOEsaK8oH8+R3ZYzLSeT5OWNQqcQXPUHoqcrLy0lPT0elUvHEE08QCoVEqPcwXRbsFosFt/tgbU6W5WioL1u2jAMHDvDFF18A8Pvf/56ioiIGDx78k/tsX+sXukZxcXGny71yI7v9kcka4tR9SaaQ8l31lFN/yLqyEiYkBzCp4sniN/iqJTZXbwMgIHvY5f8UmSAx6j4s25LJP9aW0cem446iGLZsXN91J9eD/Nh1Fo4ecY2Pvn//+98sWbKEp59+mpSUlOg47+Ja9yxdFuxFRUUsXbqUM888k/Xr15Ofnx99z263YzAY0Ol0SJKE1Wqlubn5Z/dZWFgoGmV0oeLiYoYPH95hmaIoSJKEoijo9nrIiiskyf7T386D4QBbVr6DwxbLiMIRAITlEGqVJjIS1Z46EizpbKhJ5OF1y0ixGfniD2eQFXtiDC/Z2XUWji5xjbvGGWecwb59+ygoKKCpqUlc4y7k9/uPuDLbZcE+adIkvv32Wy688EIUReHee+/lueeeIzMzk9NPP50VK1Zw/vnno1KpKCoqYsyYMV1VFOEIhOQg68s+xx90M6bveUiSxMicGYe1bVn9FgBsxngqnbvYXLEMSZKYNPByJElidN65LC85wO/++xlmnYYPrphwwoS6IBxPmpqaeOyxx/jzn/+MTqdj9uzZzJo1C7VaLWrpPViXBbtKpeKuu+7qsCw3Nzf6+w033MANN9zQVYcXfoW6lv0s3/UmjZ4DWA2x+ENe9JrDG5tdUWQ2lS8FJKobd7O9agUAyfYcQnIQjUrL9pomzlm4lLCs8N7l4xmaFtuFZyMIwpH6+9//zmOPPUZiYiJXXnklkiShVqu7u1jCzxAD1AhRshJmbeknbCr/GgWZgpRRDM8+A6368B9/7G/YRqMnMopck7eW7PjBFKaNI96aDkB1s5dpz3yB0xtg4YWjmdwvtUvORRCEI+Pz+TAYDADcdNNNJCcnc+mll3ZvoYRfRIzzJwCRUN/j/5KN5Usx6+1MKbySU3LP/kWhrigKG/YvBUAlqZk25Dp+U/DbaKi7/EFmPPslpQ1u/veMIcw9KfendicIwjG2YsUKRowYwdKlkf/HNpuNa6+9Fq1WzNNwPBE1dgGIBLFNnUZWfAEj+kz9RYHeprqphHpXOSn2XPKTT44GepsnV+xkbXkDl4/MY8HEQUer6IIgHCUWi4XGxkb27NnDaaed1t3FEY6QCPYTWL2rgq2V3zKm7yxUkppETX9G5I044v1t2P8lAEXZZ5BgjUz+EpZlttY08f2+Ot7euA+Ay0bmikGJBKGHWLp0Kf379yc5OZnBgwezYcMG4uLiurtYwq8ggv0EFJZDbNy/lI3lS1EUmT7xQ0iP7ferwnZLxXKqm/ZgM6bxdYnM6rJiVpXVUVzegDsQiq4XZ9KTLVrAC0KP8M033zBr1iymT5/Oiy++CCBCvRcQwX6CqXdVsnzXmzjdVZj1dsbknRedL/2XavIGWL2/nuKyMuJ1H6JWwWMrZL4uXQaAJMGAJDsnZcQzMiuekRnxFKY40KpF0w5B6AnGjBnD5Zdfzu9+97vuLopwFIlgP4FsrfyW1Xs/RFFk+iadxEl9pqHTGA5r20AozMaqRlaX1bGq9Wf7gWZUksKC8SWoDQpOr45Y6xDumxbPSZnxjEiPw2oQjW4EoadwuVzceeedZGdnc/3116NSqXjooYe6u1jCUSaC/QRi0Tswaq2M6TuLtJj8Du8pisJOpw+5rC7yGthd1xIN8nUVDfhDcnR9q17LhLxkzuhbQZzBB8C5Q2fzp0miUZwg9FR+v5/FixeTkZHBNddcI/qk91Ii2E8AbcPCZsYNJNWRj0bdsRYdCstc+NI3vLupDD4uOWR7jUpicGpM5JZ6ZjwjM+Pol2jD6ankg/WR8f5thniy4gcek/MRBOHwuVwuqqurycvLIy4ujnfffZfc3FwR6r2YCPZerrxhO1sqljMq7xxsxvhDQl2WFa54YyXvbiqjMM7IlEEHx4FPtRkZmRnPsPRYgqEm9jdsZX/DF/SJnYNapcJhSkJCQkFhcMZpSJJ4di4IPYnb7Wb8+PGo1Wq+/vprjEYjAwYM6O5iCV1MBHsvFpKDfF+yGJfPSVgOHfK+oijc+N5qXlpTwsmZ8dx/chzjThne+p5Mbct+9jes5ZNNW6OjyQEcaCkjK24gwZAfSVJj0lnISRh6zM5LEITDYzabOfPMM9FqtahU4ov3iUIEey+2pXwZLb56BqSOIcacfMj7dyxZz7+/3cGgFAfv/f5UNm35HF/QjUFrxhd089HGJwEFtUpDRmx/MmL7E2dJp7qphLX7PqHBVYmshChMG4dKJW7rCUJPsHLlSj7//HNuv/12AO666y4xbsQJRgR7L+XyOdlY/hUGrYWhmZMOef+hpVu49/PN5MVbeeOSQpbveJrmYC21zf3IiBuAUWelKGsyMaYkkuw5HGjex87q71m5+z0UDjaiM+qs9E068kFtBEE4ehRF4Y477mDdunXMnj2bgoICEeonIBHsvdSqvR8QloOMzjv3kC5tT63cyV8+WEu63cjC82x8t/sZZCVMrDoXR7uafW7iMHbVrOG7kvdx+xsBiDWnkp88khhTZD2rMQ6NWnfsTkwQhEM0NDQQGxuLJEk89thjNDU1UVBQ0N3FErqJCPZeyONvprpxD4m2bHIShnV479W1e7nu7e/JilHz8LRGdtWswaA1c2rf2dTsdWPW2ymr38LO6tVUOHegoKBR68hPHkl+8kjiLek/clRBELrD/fffz5NPPsk333xDRkYG/fr16+4iCd1MBHsvZNLbOHf4TQTD/g634RZv2c/cV7/FptfyzHkZlNUtJsWex9h+5xOWQ2wILubN1UvwBloAiLdkkJ88kj4Jg49oUhhBELpeZmYmCQkJNDY2kpGR0d3FEXoAEey9TFufdaPOgpGDY7J/uauKOS99jUUHi6+YwJg+iZTW2VAUheU736SycTegoFUbKEgZRX7ySGLNKd13IoIgdMrr9fLss89y1VVXodVqmTNnDjNnzozOoS4IIth7EZevkc+2LGREnzPJiD34fO37fbX87pXPuHFUKYNSMxiUrGLN3o/ZfaAYX9AFQKItG503gd+MmCGemQtCD/bQQw/x8MMPo1KpuPbaa5EkSYS60IEI9l5kTemHNHkPRMMaYGOlk2veWMxNY0pwGEMocjnvFEfGhtZrTAxIPZX85JNwmJIoLi4WoS4IPVA4HI6OFHfDDTegUqm49NJLu7dQQo8lgr2XqGzcRWndJhKsmeQlFgGwq7aZWxe/zrUnl6JpHZvCF3STbM8hP/lksuIGolaJPwFB6Mk2btzIVVddxf/93/8xbtw47HY7CxYs6O5iCT2Y+FTvBcJyiO/3vI+ExCm5ZyNJKvY73fxr6dPMHlgXXS83sYghGROwGeO7sbSCIPwSoVCIPXv2sHr1asaNG9fdxRGOAyLYe4Ftld/S5K2lIOUU4ixpVDQ6ue+TFxmeGgn1WHMao/POJd4quqoJwvFg3bp1pKWlkZiYSFFREcXFxaLFu3DYRLAf5xRFoax+K3qNmSGZE/luzyesK1vGyPQwgbCR8X1nkJs4TIw+JQjHiVWrVjFt2jSmTp3Kiy++CCBCXfhFRLAf5yRJ4ozB89hVvYb31z2KN9CEVgVlLQP4n8kXisZwgnCcGTFiBLNmzWLOnDndXRThOCWC/TgmKzLN3jpWl3xAReNOFAVkBXY5C7n3rN+iFrM5CUKPFwgEeOihh0hISODKK69EpVLx5JNPdnexhOOYCPbjlMffzPvrH8EXdEeXNfk1rK0+iWfnzBChLgjHiaamJp577jliY2O57LLL0GjEx7Lw64i/oOOMLIfZXrWS4n2fEJaDgAQo7KwzsuHAcN6+bCpatQh1QejJgsEgNTU1pKenk5CQwGuvvUbfvn1FqAtHhfgrOk4oikK5czur935Es7c2urzBl8rGKg+7nYV8PG8yBq2YF10QejKfz8fUqVPxer189dVXGAwGhg8f3t3FEnoREezHAae7mtV7P6SycReRGnqEO5jCgk+t9EvM5strJ2I1aLuvkIIgHBaDwcDIkSPxeDyEQqHuLo7QC4lg78F8QRfr9n3GzupVKCgk2fpQ21KGrITxhvtwwwdG8uJtLJk3kViTmH1NEHqqrVu3smTJEubPnw/Afffdh0q0gxG6iAj2Higsh9hW+S0b9n9JMOzHoo/hlNyz0aoNfLL5Pyhoue8rFel2M59eNZFkm7G7iywIwo9QFIU//elPrF69mkmTJjFo0CAR6kKXEsHeg0QGm9nCmtKPaPE1oFHp0Kgi/dDNegefbP4PsiLz6MoMArKdr6+bSFas5Wf2KghCd3C73ZjNZiRJ4uGHH6a8vJxBgwZ1d7GEE4D42thD1LsqWLLpaZZuf5kWnxOTzk5IDgAKeYkj+Gr7K/iCbtbX9GdDtZnHZp5Mv0R7dxdbEIROPPXUUxQVFVFRUQHAgAEDmDx5cjeXSjhRiBp7N/MEmlm371N21RQDCjZDPC6/E0+gifSYfpycezb76jbR5K3Fairi0RV+xuYkMmtwZncXXRCEH2E0GlGpVOzfv5+0tLTuLo5wghHB3k1C4SBbK5ezsXwpoXAAhymJEVlTWV36IXqtiZNzziIrrhBJkhiYNg6D1sb5L5UjSX7+ftYIMfa7IPQg4XCYV199lQsvvBCNRsMll1zCOeecg81m6+6iCScgEezdoKaplGU7X8Ptb0SnNlGQMoKROf+fvfsOj7LMHj7+nZreExJSaIEAElpAijQVsICKgpQFwYaCLvoqrIqIyLKANMtv7YiiIkpRLEhRAghLkRKaIBAglISEJKTXqc/7x5CBIQlBJZlkcj7X5XVlnjJzZiQ5c7dzD0St0uDtEYin3he91p3swlQCvcNRqVT8etqbg2m5jOncjE5RQc5+C0KIK7zxxhvMmTOHnJwcnnnmGVQqlSR14TQyxl7DFEVhx8lvKTLYutpVKjiRvociQx4A/p4N0GvdSco8yI8H/svhlK3klxp5dd0BPPUaZg3o6OR3IIQA2+9ymSeffJLHH3+ckSNHOjEiIWwksdewjPyz5JVk4q7zIiXnOGariY6N+uGlvzwRLj3/DNsSV6LTuBEREMOcjYfJKCzlpdtjCffzdGL0QgiAU6dOMSH5ZIQAACAASURBVHDgQLZv3w6Av78/8+fPJyhIetOE80lir0GKYmVX0g+ArfhMZEBL7o97ntjIPqjVtlKw+SUX2fTHFyiKldtaPURuqRdvbz1KlL8nE/vc5MzwhRCXZGdns3v3btavX+/sUIQoR8bYa5DRYiC76AIqVPRuOYImwe0cJsEZTMXE//EZBnMxtzQfTHhAC4Z/sRWD2crsgXF46uV/lxDOkpSUhK+vL8HBwdx8881s3bqVm26SL9ui9pEWezUzW01cLEyh2JBPUsYBQKFd1O00DWlfbmb7H6nbyC+5SGxEH2LCurAtKYNvDp6la6Ng/tGxiVPiF0LAoUOH6NWrFy+++KL9mCR1UVtJE7CabUtcwZmLv9M4qC1ZRedRqdS0bNitwmvbN+qLt1sAzUM7YbUqTPxhDwBv3i/L24RwpjZt2nDbbbdxzz33ODsUIaokib0a5Ranc+bi7/h6hKDT6CkszaZRUBs89T4O1+UVZ+LnGYJapaFF2M0ALEk4RUJKNiM6NqFb4xBnhC9EvWW1Wvnkk0/Q6/U8/PDDaDQavvzyS2eHJcR1qbIr3mg08sEHH/Diiy9SWFjIu+++i9ForInY6rxDyb8C0LnJ3ahVtslxMaFdHK45lbGf7/a9SeKF3fZjRQYTr6zZj7tWw+sD42osXiGETXZ2NnPmzGHBggUYDAZnhyPEn1JlYp8xYwYlJSX88ccfaDQazp07x5QpU2oitjqtoDSL05kH8fcMJcyvKUkXD+Dl5k94QAv7Nel5p9l+4ht0Gjca+Da2H5+/+Q9S80uYdOtNNArwckb4QtQ7VquVjIwMAIKDg/niiy+Ij4/HzU22RBZ1S5WJ/ciRI0ycOBGtVouHhwdz587l2LFjNRFbnXYu6w8UrLSLuo0zF3/HbDHSIrQzapXtI88vucimo0tQULit9UP4e4YCkJxTxIJfj9DQ14MXb2/jzLcgRL1hMpkYMmQIDzzwgL2F3qNHD0JDQ50cmRB/XpVj7CqVCqPRaJ+8lZOTIxO5rkObiF6E+jYh0DuctQc/RIWKFqG28fNSUxEbjizGYC6mR/MhhPs3t983Ze1+SkwW3h3cEW83nbPCF6Je0el0NGvWDA8PD4qLi6WVLuq0KhP7mDFjePTRR8nMzGTWrFnEx8fz9NNP10RsdV6wTxTZhalcLEwmMqAVXm626nJ7T6+loDSLdpG32SfLAew6m8lX+04TFxnImM7NnBW2EPVCcnIya9euZdy4cQC8/vrr6HQ6abiIOq/KxH7//fcTGxvLrl27sFgsfPDBB7Rq1aomYquTSk1F7Dm9htiIPgR4hZKYbluyFnNFAu8WfT/hAS1oGtzOfkxRFCb9kADAG/d1Rq2WPy5CVKfx48ezc+dOOnfuTKdOndDr9c4OSYgbosrE/swzz/DOO+/QvPnl7uKHH36Yzz//vFoDq6uOpu7gVMY+grzC8XEP4FTGfjz0PkQGXv4ypNXoaBbSweG+5QfOsPNsJoPbNaJ3tIzrCVEdTCYTOp1tiGvu3LkcOnSIuDhZeSJcS6WJfcKECRw9epSMjAz69u1rP26xWAgLC6uR4Ooak9nA0dTtuGm9aBHWhTMXf8dkKaV1w+6oVRqKDHmczTpMi9Cb0Wkutw5KTGZeXrMfvUbN3Hvkj4wQ1WHFihXMmjWLdevWER4eTmxsLLGxsc4OS4gbrtLEPmfOHHJzc5k1axZTp069fINWKzsYVeL4hd8wWkrp2PgOdBq9vRu+RVhnAP5I3c6R81vRqfUOY+tvbTnKuZwiXritDc2CfCp8biHE31NUVERubi7Hjh0jPDzc2eEIUW0qXe7m7e1NZGQkH3zwAXl5eaSlpZGamsrp06f5/vvvq3xiq9XKtGnTGD58OKNHj+bs2bMO57ds2cKwYcMYNmwY06dPd9jbuC4yW00cPv8/dBo3WjXsTm5xBhn5Z2jo3xwf9yAM5mKOX/gND70PzRpc3lM9Lb+YORsPE+LtxpR+0noQ4kZRFIU1a9ZgNpsBeOSRR9i9eze33367kyMTonpVuY596tSpPPfcc/zzn//kzTff5KmnnmLdunVVPnF8fDxGo5Hly5czadIk5syZYz9XWFjI/Pnz+fDDD1mxYgURERHk5OT8vXfiZCfT91JqKqRVw+64aT04camSXFmlueNpuzBbjLQJ74lGfbmjZOraAxQZzcy4qwO+7jJ5R4gb5b333mP06NF88MEHgG3prqxLF/VBlYl9x44drFmzhjvvvJP//Oc/fPHFF5SWllb5xAkJCfTq1QuADh06cPjwYfu5/fv3ExMTw9y5cxk5ciTBwcEEBgb+jbfhfJEBrWjd8BZuCu+JxWrmZMY+3LReNAq6CbPFxB+p29Fp3IkJ62q/Z19KFp/vPUXbhv483rX5NZ5dCPFnjRw50l50Roj6pMpZ8Q0aNECn0xEdHc3x48cZOHAgBQUFVT5xYWEh3t7e9scajQaz2YxWqyUnJ4ddu3bx/fff4+npyahRo+jQoQNNmza95nNe+eWgNtISwR+5x8k1J2MwFxOsjeHA/oNkmU9RaiokRNuK3w8eAWzdhE9tPIuiwLjWvhzYv9/J0dskJCQ4O4R6QT7nG+/ixYu88847DB06lNjYWE6fPs348eNJT08nPT3d2eG5JPl3XDtVmdhDQ0P56KOP6N69O/Pnzwe4rk1gvL29KSoqsj+2Wq1otbaX8/f3p23btoSE2HYt69y5M0ePHq0yscfGxta6ilCKYuVC3mnC/JrZC1v8/Pt+yIOe7e7B37MBqTm+WJKz6dNqiH1nt1WHzrEv4yj33BTJUwP7OPMt2CUkJNCpUydnh+Hy5HOuHr/99hs7d+4kOjqa2NhY+Yyrmfw7rl4Gg+EvN2ar7IqfNWsWkZGRtGvXjjvuuIOffvqJf//731U+cVxcHFu3bgXgwIEDxMTE2M/FxsaSmJhIdnY2ZrOZgwcPOqyTr0vOZh3h58Mfc+BcPAD5JVmk5Z0k1Lcp/p4NAAgPaMHd7cbbk7rBbOGlnxLQqlXMv09+MYT4q9LT0+3zc7p168batWt56623nByVEM51zcReVFSEm5sbAwcOBGD06NG8/fbbHDhwoMon7t+/P3q9nhEjRvD666/z8ssvs3jxYjZu3EhgYCCTJk1i7NixDBs2jP79+zsk/rpCURR+T/4VUNGsga3gzAl7pbkuKIpCibGw3H3v/O8YSVmF/LNnS2JCfGswYiFcx7Fjx7jlllt48cUX7ce6du2KWl1le0UIl1ZpV/yyZcuYOXMmnp6eLF68mDZt2rB+/XrmzZuHp6envb5yZdRqNTNmzHA4Fh0dbf954MCB9i8MddX53ESyis7TJLgdfh4hWK0WTqYnoNe40zgolrTck8T/8Tndou8jJsw2Oz6joIRZ8b8T6Knn1f7tqngFIURlWrRoQbt27ejatSuKokiNdyEuqTSxL1q0iG+++YaUlBQWLlyIr68vmzZt4plnnmHo0KE1GWOt9XvyZgDaRt4KQHLOMUpMBbRueAtajY7fU7ZgVcwEel8uhvHazwfJLzXx3wduJsCzds0XEKI2UxSF7777DqPRyIgRI9BoNKxatUoSuhBXqTSxe3h40KpVK1q1asXUqVPp3r07P//8s8NM9/osPe806flniAxoRdClxG1fux7WhYuFKaTlnaShX3OCvSMB+D0th0W/naRVA1+e7F73hh6EcKbc3FwmTZqEXq9n0KBBeHh4SFIXogKVJnaNRmP/2c/Pj3nz5tk3TxBQYirEXedFu6jbACgszeV8TiLBPlEEeIXx67GlALSNtM14t+3etherojD/vs7oNDIOKMT1yM/Px9fXl4CAABYtWkTTpk3x8PBwdlhC1FqVJvYrvwl7enpKUr9Kk+C2RAa2Qqu2fS4n0vegoBAT2oX8koucuXiYIK8IGvrbZvuvOXqejScucEfLcO5uJXWqhaiK2Wxm3LhxnDhxgvj4ePR6vcOGVEKIilWa2M+cOcOYMWPK/Vzmiy++qN7I6oCypG6yGDiRvhetRk/TkHacytgPKMRG9kGlUmGyWHnhxwQ0ahUL7usk3YdCXAetVou3tzeenp7k5ORIOVghrlOlif2jjz6qyTjqDIvVzLrfP6JZcHtuiuiJoijsOLGKYmMesZF9Lm0C040wv2b4egQD8MH24yRm5vPULTG0CfN38jsQovbKzs5m3bp1jBo1CrDtMqnX6x2GBoUQ11ZpYu/SpUtNxlFnpOef5mJBMg18GgFwLG0Hpy8epIFPYzo26m+/rqw4TXaxgRm/HMLPXcf0O9s7JWYh6orHH3+cLVu20Lx5c7p27Spj6UL8BVWWlBWOUrKPARAZ2IqM/LPsPr0Gd50XfVqNRIWKrceX0Tq8ByE+UQDM+OUQOSVG5t/biWBvd2eGLkStZLVa7UVlpk+fzpYtW+jcubOToxKi7pKp2X9SSvZxtBo9vu4htpnvikKflv/Ay82PjIKzJGUe4FTGPgCOpefx/vbjNA/2YULPlk6OXIja55dffqFnz572TVrat2/Ps88+K13vQvwN15XYU1JS+PXXX7FYLCQnJ1d3TLVWfslF8ksv0tAvmm0nVlBszCeuyZ32me8pOccBiAywJfEXVidgsSrMvScOvVb+UAlxtbNnz5KUlMSePXucHYoQLqPKxL527VqeeuopZs6cSW5uLiNGjOCHH36oidhqneRL3fBWxcKFvCSiAm8iNuLyzmzns4+jUWsJ82vGL8dTWXv0PLc1D2VQbJSzQhai1tm+fTsWiwWwjanv2LGDe+65x8lRCeE6qkzsH3/8MV9//TXe3t4EBQXx3XffsXDhwpqIrdYJ8g4nzK8Z53MS8XEPolfMMPvStSJDHjnFFwjzawZo+dePe1GpYMF9nWV5mxCXfPbZZ9x77718+OGHgG1PiWbNmjk5KiFcS5WJXa1WO5SRbdCgQb3dPclT70tW4Xk0ah23tX4IvfbyZLjzV3TDf7zrBEcu5PHozc3pEBHorHCFqHXuvfde+vbty2233ebsUIRwWVVm6BYtWvDll19iNps5evQor776Kq1ataqJ2GoVk9nA5qNfYrIYuKX5AwR6NXQ4r1Zp8PMIwcejGdPXH8TbTct/7u7gpGiFqB3y8vJ49tln7WPoQUFBrFy5kptuusnJkQnhuqpM7NOmTSM9PR03NzemTJmCt7c3r732Wk3EVmsoisLqg++QU3yB6JA4ohvElbumeWgnHug0if9uS+VikYGX+8YS5itrcEX99scff/Dll1/ywQcfODsUIeqNKtexr1y5kkceeYRJkybVRDy10vG0XeSXXESFmm7N77/mtZ/tPkWojzvP9ZYWiaif8vPzAfD19aV79+6sWLGCPn36VHGXEOJGqbLFfuHCBYYOHcrYsWP58ccfKSkpqYm4ao2LBcnsSvoRgKjA1ug0+nLXbD66hP1nN6AoCgaLhXBfT9x1srxN1D+nT5+mR48evPTSS/Zj/fr1k02khKhBVSb2l156iU2bNjF+/HgOHDjA/fffz4svvlgTsTldqamIzceWomAFoElIu3LXZBelcTbrCBcLUziTXUihwUyUv2dNhypErRAVFUV4eDiNGjVCURRnhyNEvXRdJWUVRcFkMmEymVCpVPXi27dVsbL1+DKKDLl46f0oNuYTEdCi3HWJF3YB0DKsC+sT0wDoHyPbsor6Y8uWLWRlZTF48GC0Wi1r1qxBq5Vq1UI4S5W/fTNnzmTDhg20bt2a++67j6lTp+Lm5lYTsTnVwXMbSc09QUO/aNLykmjg2xg3rWNL3GQxcipjP556XyIDWxGfuA2A/i0bVvSUQric/Px8HnnkEVQqFXfccQfe3t6S1IVwsip/Axs3bsx3331HYGD9WY+dkn2Mg8kb8XYL4JbmQ0jJOY6H3rvcdaczD2KyGLgpvCeKomLTiQs0DvCiebCPE6IWouaUlpbi7u6Or68v7777LuHh4Q71LoQQzlNpYl++fDnDhw8nLy+Pr776qtz5CRMmVGtgzlJQms3WxOWoVVpua/0QPh6BtPboXuG1iRd2o0JFi9CbSUjJJqfEyJD2jaTSnHBZVquVyZMns3fvXn7++Wd0Oh0DBw50dlhCiCtUOnmuPk58MVtN/HpsKUZzCd2iBxHo1RCr1VLhtYqiENfkTuKa3Im3uz8bjqcC0E/G14ULU6vVFBUVUVxcbN+RTQhRu1TaYh8xYgQAERERPPDAAw7nli5dWr1ROcnuU6vJKjxPi9DOxITdTHreGeL/WEznpgNoGdbV4VqVSkW4f3PCL+3sFp+YhkoFfVuEOSN0IapNUVERGzdu5L777gNgzpw56HQ63N3dq7hTCOEMlSb2zz77jMLCQpYtW8b58+ftxy0WC6tXr2bUqFE1EmBNOZG+l8T03QR6hdO12SAAUnKOYbIY8NQ5jpmbLEZKTQX4uAcBUFBqYseZTDpHBhHo6foTC0X98sQTT7B+/XrWrFlD9+7d8fGROSRC1GaVdsU3adKkwuN6vZ45c+ZUVzxOkVV4nt9OfY9e485trUeh1diW86VkH0Ot0hJ2qVVeJilzP9/unc+pjH0AbElKx2xVZDa8cEn/+te/eOaZZ+jQQfY+EKIuqLTFfuutt3Lrrbdy9913Ex0dDUBhYSFpaWm0aFF+PXddZTAXs/noUixWM7feNMreCi8szSWn+AIRATHlqs0lptkmzYX52T4XGV8XrmTXrl289tprLFmyhJCQEOLi4oiLK78/ghCidqqy8ty+ffuYPHky2dnZDBgwgGeffda+l3JdpyhW/nd8BYWGbNpF3U5UYGv7ucvbsDruZHexMIWsovNEBrbCy80PsI2ve+m1dG8cXHPBC1FN9u/fz549e9i0aZOzQxFC/AVVJvavv/6aiRMn8tNPP9G3b19Wr17NL7/8UhOxVbvfU7aQknOMcP8WdGjUz+FcSs4xACIDWzocT0zbDWCfTJecU8SxjHz6RIei10p9eFE3HT58GKvVVjr5ySefZNOmTQwfPtzJUQkh/ooqEztAgwYN2LJlC7feeitarRaDwVDdcVW71NwT7D/7C15ufvRuOQK1yvGj6NCoH12a3mPvmgcwmktJyjyAl5s/4QExAGy4VEb2DhlfF3XUypUrufXWW/n4448B25K29u3bOzkqIcRfVWXluebNmzNu3DhSUlLo3r07zz33HO3ald8MpS4pNuaz5dgyVCo1t7YahbvOq9w1Qd4RBHlHOBy7WJCMVbEQE3qz/YvAhkQZXxd1W58+fejYsWOd/70WQthUmdhnz57N/v37iYmJQa/Xc99999G7d++aiK3apGQfw2AuomPjOwjxaVTufImxEHedV7kKcuEBLRjW5WVUl5K61aqwMfECkX6etGrgWyOxC/F3GQwG5s6dyz333ENcXBwNGjTgl19+kYqJQriIKrviTSYTmzdv5tFHH2XQoEH89ttvGI3Gmoit2lgV21jild3sZRRFYe2h9/lx//9VWH3PXedt3wxm//lssooN9ItpKH8URZ2xf/9+3n77bebNm2c/Jv9+hXAdVSb2GTNmUFpayuzZs5k7dy5ms5nXXnutJmKrNopiKxNb0Z+yvJJMCkqz8fUIdvhjd+T8Nk5nHrJ/KQDbbHiQ3dxE7WcwGCgsLASgW7dufPrppyxatMjJUQkhqkOVXfFHjhzhxx9/tD+eNm0aAwYMqNagqlteSSZQcYs9JbtsNvzlZW5Gcyn7z/6Mu86HJsGx9uNl4+t9W0hiF7VXSkoKw4YNIy4ujnfffReA+++/38lRCSGqS5UtdkVRyM/Ptz/Oz89Ho6nby7qyC9NQocbfK7TcubL16xEBl5e5ncrYj9lqIibsZvv4epHBxPbTmcRFBhLiLTWzRe0VGhqKm5sbbm5uWCwVb2okhHAdVbbYH3nkER588EFuv/12ADZt2sSTTz5Z7YFVF6tiJbsoDT/PELRqncM5o7mUC/mnCfKOwFNvq4etKAqJF3ahUqlpHtrZfu3WpAyMFiv9pLUuaqFDhw5x/vx57r77bnQ6HWvXrsXDw8PZYQkhakCViX3IkCG0bduWPXv2YLVaeeedd2jZsmVVt9VaBaVZmK1GgrzKL09Lyz2Jolgdqs1lFpwjp/gCjYPa2pM9yPi6qL2Kiop44IEHsFqtHDx4EF9fX0nqQtQjlSZ2q9XKN998Q2JiInFxcS6zm1t2oW1cPNC7fGKPCGxJvzaP4ut+uTTs8Qu7AGjZsIvDtRsSU/HQaejRtEE1RivE9bNYLGg0Gry8vJg/fz4BAQH4+soyTCHqm0rH2KdPn84333yDTqfjww8/tE+6qeuyi2wt7UCv8i1trVpHZEBLfD0uT6oL8WlEREBLGl7a8AUgNa+YIxfy6B0dipuUkRVOpigK8+bNY+DAgZjNZgAGDx7Mbbfd5uTIhBDOUGmLfc+ePaxduxaVSkVOTg4PP/wwEyZMqMnYqkVWWYv9qq74EmMhimLF082xhdOqYTdaNezmcKysjGz/GOmGF86nUqlISkoiJSWFc+fO0axZM2eHJIRwokpb7G5ubvZ13AEBAS5TwCK7KBUvN3/cdJ4Ox49f+I0Ve2aTcmlWvKIoWKzmCp8jXhK7cDKTyeSw+9rcuXPZvn27JHUhROWJ/epErlZf134xtVqxMZ9SU2G51jpASvZxVKjtJWYzCs6yYvfrJGUccLjOalWIT0yjoa8HbcL8ayRuIa729NNPM3ToUHbu3AmAn58ffn5+To5KCFEbVNoVn5qayssvv1zp49dff716I6sG2YUVj6+XGAu5WJhCqG8T3LS22cOJabswmIvwuGImPMDvF3LIKCxldOdmLtOLIeqeJ554Ar1eT+vWrZ0dihCilqk0sU+ePNnhcZcuXSq5su7ILrKNrwddNSPeVpRGIepStTmDqZjTF3/H1yOYMD/Hrs0Nx21fDvpJN7yoQceOHeO1117j/fffJygoiC5durjE76QQ4sarNLE/8MADNRlHjShL7Fd3xafYq83ZEvvJjH1YFTMxoV3KtcrLJs5JYRpRkzZu3MiGDRv44YcfeOyxx5wdjhCiFquyQI0rySpMRa/1wMvt8ti4VbGSlnsSLzd//D0b2CvNqVUamod2cri/xGTmf0nptGsYQJivFPwQ1evMmTM0atQItVrN+PHjadeuHb169XJ2WEKIWq7uz4i7TiazgYLSLAK9HLdYVavUPNBpIre2GoVKpSKzIJm8kkwaB8firvNyeI5tSRkYzFapNieq3dq1a+nevTuLFy8GQKPRSFIXQlyX62qxFxcXc+7cOVq2bElJSQmenp5V31TLlBWmqaiUrLvOG3edNwAhPlHc3W48ek35Frm9G17G10U1i4uLo3HjxoSHl//3KoQQ11Jli33nzp0MGjSIp59+mosXL3Lbbbexbdu2mojthrKPr181ca7IkEepqcj+WKVSEerbhIAKdn6LT0zDTaumVzMpIytuLIvFwnvvvcfBgwcBCAsLY8eOHdx9991OjkwIUddUmdjffPNNvvrqK3x9fQkJCWHp0qXMmzevJmK7oSqbOLf/7C8s2/UfcoszSMs9RW5xRoX3pxeUcDA1h55NG+Chq1dTE0QNSEhI4NVXX2X69On2Y65QO0IIUfOq/MthtVoJCQmxP27evHm1BlRdsgvTUKu0+HmEOB4vSkOj1uLjHsjOU9+x+sB/MZkN5e4vqzZ3R0vpGhU3htVqpaSkBLAtJ/2///s/Pv74YydHJYSo66pM7GFhYWzevBmVSkV+fj4ffPDBdY37Wa1Wpk2bxvDhwxk9ejRnz56t8JqxY8fy9ddf/7Xor5PVaiGn+AIBXmGo1Zc3bbFYzeQWpxPgGUZG/lnySy7SJKgtOq1bueeQ8XVxI2VkZHDPPfc4FH0aPXo0wcHB17hLCCGqVmVinzFjBqtXryYtLY1+/fpx9OhRZsyYUeUTx8fHYzQaWb58OZMmTWLOnDnlrnn77bfJy8v7a5H/CbklGVgVS7mKc3klmZeOh9u3Z40J61rufkWxlZFt4O1Ou4YB1R6vcH3+/v4UFBSQm5tr35FNCCFuhCoHi4OCgnjzzTf/9BMnJCTYl+d06NCBw4cPO5xfv349KpWK3r17/+nn/rPK9mC/uuJc2XEf90D2n9uAv2cDGvg2Lnf/kQu5pOWXMDKuKWq1lJEVf82ZM2dISkrCz88PvV7PTz/9hK+vr5QmFkLcUFUm9ttvv73CPzwbN2685n2FhYV4e3vbH2s0GsxmM1qtlsTERH766Sf++9//8t577113sFd/ObheqUbbRi4ZyfkUpSaUO346JRGrYsHDFM6+ffvK3f/VsSwAmusNJCQklDvvSlz9/TmL0WjkoYcewmw289lnn8nnXAPkM65+8hnXTlUm9iVLlth/NpvNbNiwAaPRWOUTe3t7U1R0eRmZ1WpFq7W93Pfff096ejoPP/ww58+fR6fTERERUWXrPTY2Fje38uPfVck4tBfyVXTvdCs6zeX7WxmiySxI5mJBCkXpF7i10324acuv0X91n+1LzBN3dCfcr+6t4b9eCQkJdOrUqeoLxXVTFMX+xXjmzJnodDp8fHzkc65m8m+5+slnXL0MBsNfbsxWmdgjIiIcHo8dO5bBgwfz9NNPX/O+uLg4Nm/ezIABAzhw4AAxMTH2cy+++KL953feeYfg4OBq7ZLPKb6Ar3uQQ1IH8HLzx8vNnybBbenYuD8adfmPw2C2sPVUOm3C/Fw6qYsbS1EUFi9ezOrVq/nmm2/QaDT84x//AKSVI4SoXlUm9j179th/VhSFEydOYDCUXw52tf79+7N9+3ZGjBiBoijMnj2bxYsX06hRI/r27fv3ov4TTGYDRnMJwd5RDsctVjMKClq1DqDCpA6w/XQGJSYL/WNkmZu4fiqVip07d3Lw4EESExNle1UhRI2pMrH/97//tf+sUqkICAiocIb71dRqdbnZ89HR0eWue+aZZ64nzr+syGibde/lqFR+ggAAIABJREFU5udwPDX3BBv/+AIf90C6NruXyEtbtl4tXpa5ieukKAp79+7l5ptvBmDu3LkYDAYaNpR/O0KImlNlYh8wYIC9C7EuKjZUnNhtM+IVCkqzKCjNqfT+DYlp6DVqeksZWVGFiRMn8sUXX7B27Vq6du1KYGCgs0MSQtRDVa5jX7p0aU3EUW2KDLkAeOkdE3vWpaVuGrWO6AYdK7w3s7CU/eez6dE0BC83XfUGKuq84cOHc8cddxAVFVX1xUIIUU2qbLGHhYUxZswY2rdv7zAjfcKECdUa2I1S1hXveVWLPbPgHABNg9uj17pXeO/GE2koCjK+LiqUkpLCf/7zH+bMmUNAQADdunWjW7duzg5LCFHPVZnYO3ToUBNxVJsie1e8v/2Y0VxKiakAgJYNy1eaKyPj6+JavvvuO1auXEmbNm149tlnnR2OEEIA10js3333HQ888ECdaZlXprhs8twVXfFpuacAcNd6EewdWeF9iqKw4XgaQZ5udIyQsVJhk5GRQUhICCqViqeeeorGjRtz7733OjssIYSwq3SM/YsvvqjJOKpNkSEPncbdYWMXf68wmoV04KaInpWW8zyekU9KXjF9Y8KkjKwA4Ndff6VLly58/vnnAGi1Wu677z4pCSuEqFVcfmPxIkOuQzc8gJ9HEL1bjrjmfRsSbZPrZHxdlImJicHHx+cvVT8UQoiaUmliP3HiRIWFZMpKZFZVK742MJpLMVkM5Za6Wa1W1OprLwgo26a1v4yv11uKorBixQratGlDbGws4eHhJCQkoNfrnR2aEEJUqtLE3rhxYxYuXFiTsdxwl8fXL7fYS4wFLN89iwDPMAbFPVfhfUazhV9PptOqgS9RAV41Equoffbv389TTz1F9+7dWbNmDYAkdSFErVdpYi/bmKUuK5sR7+nmaz+WnH3U9sM1xkV/O3uRIqNZZsPXQ4qiYDab0el0xMXFMXv2bAYMGODssIQQ4rpV2h8dFxdXk3FUi4qWup3POQFQ6Wx4uGJ8vaWMr9cnubm5jBkzhpdfftl+bPz48TRq1MiJUQkhxJ9TaYt92rRpNRlHtahoqVt20XkAIgJaVnpffGIaWrWKPs1CqzdAUat4eHiQlJREXl4eRqNRut2FEHWSS8+KLysnW1Z1TlEUeys+3L95hfdkFxvYk5xFr6YN8HGXMrKuLiMjg5MnT3LLLbfg5ubGqlWrCAkJqXJypRBC1FYuntjzgcsbwBQZcrEqFjRqXaVlZDeduICiSLW5+sBoNNKvXz+Kior47bffCAkJITRUemmEEHWbSyf2YmMueo07Oo1t3bFKpSYioCUeOp9K75Hx9fpDr9czceJESktLCQoKcnY4QghxQ7h0Yi8y5DlMnPNy86N/m0crvb6sjGyAh55OkVJG1hX98MMPfPvttyxevBiNRsMjjzzi7JCEEOKGctmBxMvFafyrvviSkxcLOJtTxO0twtDIGKtL+v7774mPj+fw4cPODkUIIaqFy2avq2fEWxUrX+38N9/smYvRXFrhPRuOy25urujYsWP2n+fPn8+WLVto3769EyMSQojq47KJ/fIadltizyvOxGgpociYbx9zv9rl+vCS2F3F9OnT6dmzJ3v37gUgODiYFi1aODkqIYSoPi6c2B2XuqXnnwbA282/wt24TBYrm0+m0zzYh6ZBlU+uE3VL//796dSpE35+flVfLIQQLsCFE7tjiz31UsW5oEoqzu0+d5ECg0la63VcTk4OkydPJjfX9sWuR48erF+/XlrpQoh6w2UT+9UbwGQVpgDQ0D+6wutlfN01LF26lIULF/LRRx/Zj8l+6UKI+sRll7td3gDGD7PFRJHRVqwm2LvijW3iE9PQqFXc1jysxmIUN0Z+fj4+Pj6oVCrGjx+Pr68vI0eOdHZYQgjhFC7bYi8y5KHXeqDT6LEoZiICYgjwDMPfs3xlsdwSI7vOXaRro2D8PKQ+eF2ye/duunfvzpdffgmAVqtlzJgxaLUu+51VCCGuyWX/+hUb8/B2CwDATetxzcI0m09ewKooMr5eB4WHh2MymSgoKHB2KEIIUSu4ZGK/XJzm+mZCy/h63RIfH09ERAStW7cmMjKSffv24e3t7eywhBCiVnDJxH7l+DrATwfexWQx0CtmOME+5WfFxyem4euuo0uj4BqNU/x5hw8fZtiwYXTu3Jmff/4ZlUolSV0IIa7gmondaFvq5KX3w2QxcvHSjHiTpXzFuaSsAk5lFTAoNgqtxmWnHNR5VqsVtVpNbGwsU6ZM4a677pLZ7kIIUQGXTOzFV7TY80sy7ccDvMp3tW9ItHXD928p3fC1UWFhIVOnTsXd3Z05c+YA8K9//cvJUQkhRO3lkon9cnEaf/IuJXa91h13nVe5a8vG12XiXO2k1Wr57bffcHNzo7S0FHd3d2eHJIQQtZpLJvYrN4BJybZtAOLn0aDcdWaLlc0nL9A00JtoKSNbaxQUFHDixAni4uJwd3dn5cqVhIaGotfLUkQhhKiKSyb2K8vJZhWeByouJbs3JYvcEiND2zeW8dpawmw2079/f7KystixYwchISFERUU5OywhhKgzXDSx5+Km9USr0ePtHkB2URrh/s3LXRcv4+u1jlar5bHHHiMzMxNfX19nhyOEEHWOyyV2RVEoMuTh4x4IQK+YYZVeu+F4GmqVituljKxTbdu2ja+++op3330XtVrNk08+6eyQhBCiznK59V1GSylmqxEvN/9rXpdfauS3s5ncHBVEgGfF+7OLmrFw4UJWrFhh3zNdCCHEX+dyif3KpW4XcpP4Yd/bHDm/vdx1v55Mx2xVpNqck6SkpNh/XrBgAT///DNdunRxYkRCCOEaXC6xXzlx7mzWYXKKL5Cac7zcdTK+7jxvv/02nTp1Yv/+/QA0aNCATp06OTkqIYRwDS6X2K9c6lY2Iz7Et1G56zYkpuHtpqVb45AajU9AXFwcTZo0kZUIQghRDVwusReW5gC24jQFpVkAhPu1cLjmbHYhiZn53Bodhk7KyFa74uJi5syZQ35+PgC9e/dm+/btdOjQwcmRCSGE63G5rJZXchEAH/cgSk3FAAR6hztcYy8jK+PrNeLzzz9n3rx5vP322/Zjsl+6EEJUD5f765pfkolWrUOFCgUrWo0bWo3O4RoZX69+BoMBvV6PSqVi7NixmEwmxo4d6+ywhBDC5blUi11RrOSXZuHrEUyhIRu1SoOPW4DDNRarlY0n0ojy9yQmRAqgVIcjR47Qu3dvli1bBoBOp+PZZ5/F09PTyZEJIYTrc6kWe7GxAIvVhK9HCKF+TRnTYxaKojhcs/98DtnFRu6PbSSTt6qJr68vFy5c4PTp084ORQgh6h2XSuxlW7T6egTbj12dvDccTwWQ9es3WEJCAj4+PsTExBAVFUVCQgLBwcFV3yiEEOKGcqmu+LKJc34ewew8+R27k37CarU6XBOfmIZKBX1bSBnZGyUxMZE777yTf/7zn/YeEknqQgjhHC7WYi+bER/MtsSVKECXZvfYzxcaTGw/k0lcRCDB3rKv940SExPDM888Q9++fWV4QwghnMwlE7tWpUVBwU3rOFlra1IGJouV/i3DK7pdXCej0cj8+fMxGo38+9//BuC1115zclRCCCHAxbri80su4qb1JLPwHEC5jWBkfP3GsFgs/PDDD6xevZqioiJnhyOEEOIKLpPYrVYLBaXZ+HoEk55/BoBAL8cEHp+Yhqdewy1NpIzsn2U0Gjl69CgAHh4eLF26lC1btuDl5eXkyIQQQlzJZRJ7gSEbBSu+HsHkFF0AINS3qf18Sm4Rf6Tn0btZKG5ajbPCrJOsVisDBw7k/vvvJyvLVqa3RYsW+Pj4ODkyIYQQV3OZMfZ8+4z4ENJyTwLQ0D/afj4+0Zbs75Dx9T9NrVZz3333cerUKXQ6XdU3iHrDbDaXW3lSWxiNRmeH4PLkM/771Gr1DS+x7XKJ3dcjmGFdplBYmuMwxr4hUcbX/4zDhw+zZMkSXn/9ddRqNRMmTJAZ78JBQUEBGo2mVtb9j46Orvoi8bfIZ3xjGI1GSkpKbmgPaLX9RlqtVqZPn87x48fR6/XMnDmTxo0b289/9tlnrFmzBoA+ffowYcKEv/V6VyZ2AG/3y6VkrVaFjSfSCPf14KZQv7/1OvXFvHnz+Omnnxg4cCC9e/eWpC4cmM1mNBpNrS0TbDKZ0Ov1zg7DpclnfGPo9XqKi4sxm8037EtytY2xx8fHYzQaWb58OZMmTWLOnDn2c8nJyfz4448sW7aM5cuXs23bNo4dO/a3Xq8ssZcaC/k9ZQsFpdn2cwdTc8gsNNAvpqEkqGso21YVbIl9xYoV9O7d24kRidrKarXWypa6EHWRRqO5oUNa1ZbYExIS6NWrFwAdOnTg8OHD9nNhYWEsWrQIjUaDWq3GbDbj5ub2t14vr+Qinno//kjbQcKZdfyessV+7vJubjK+XplPP/2UUaNGcejQIcD2/6hfv35OjkoIIVzfjW5wVttX7sLCQry9ve2PNRqNvatBp9MRGBiIoijMmzePm266iaZNm17j2Wyu/HJwJatiptiYh5e6Aek5tjXsJVmQkJcAwLd7zwIQUpJBQkJ2hc9R35lMJry9vdm7dy8mk8nZ4bi8hIQEZ4fwt0VHR9fqfytSY6H6yWd8Y5hMJk6dOnXDnq/aEru3t7fD//Sru+4MBgNTpkzBy8vruquWxcbGVtiyzy5M5cgBiAhpwomMvQDc3K43vh5BFBvNHFpxjA7hAfTv0fVvvivXYTabWbRoESNHjsTX15dOnTrRtm1bbrnlFmeH5vISEhLo1KmTs8P4W8pmQztzjHXXrl0899xzNG/eHLAlmcjISBYsWIDJZMJgMDB37lxSU1OxWCw0bNiQyZMnExJiq2Oxd+9e3nvvPcxmM8XFxQwePJhRo0Zd12svXbqUFStWMG7cOAYMGFDpdatWrSIpKYl//etff/8NX1JaWsoLL7xAVlYWXl5ezJ07l8DAwGve89prr3Hw4EG+//57+7HRo0czffp0+yQ4g8HA3XffzaZNmwBYvnw5P/74I2q1GpPJxPPPP0/Xrpf/hhYVFV1XHYt3332XX3/9Fa1Wy5QpU2jXrp3D+R9//JHFixejVqsZMmQII0eOtJ/Lyspi8ODBfPrppy49Wc9oNNK2bVuH3yeDwVBpY7Yq1ZbY4+Li2Lx5MwMGDODAgQPExMTYzymKwtNPP03Xrl158skn//Zr5ZfaxtfddV4oihWVSo3Ppclz/0vKwGCWMrJX++yzz5gyZQrnz5/nP//5D8DfHg4R9deLqxP45uDZG/qcD7ZvzLx7r/0FqFu3brz11lv2x5MmTWLTpk307NmTCRMm8Nhjj9mHlHbs2MG4ceNYuXIlqampzJw5k0WLFhEcHExpaSljxowhKirquuaVbNiwgXnz5tGyZcu/9yb/gq+//tq+P8OaNWt4//33mTp1aqXXl5SUsG/fPmJiYti1a5dDcq7MmjVr2L59O5999hk6nY7k5GQeeughvvvuuyq/RFzpyJEj7N69m5UrV5KWlsYzzzzDt99+63BN2URdT09PBg4cyMCBA/Hz88NkMjFt2jTc3WVfjz+r2hJ7//792b59OyNGjEBRFGbPns3ixYtp1KgRVquV3bt3YzQa+d///gfAxIkT6dix4196rbJd3ayKbfKBl94Plco2faBsfF2WudlKwarValQqFWPGjCEtLe1vr0YQorYwGo1kZGTg5+fH0aNH8fHxcZgncsstt9CoUSP27NnD3r17uf/+++27ELq7u/PJJ5+Um+WfkpLCK6+8gtlsRqVSMXXqVA4ePMjhw4d55ZVXeOutt4iKigJsLemXX36Z1NRUTCYTr776qsNzvfHGGxw+fJiioiKio6N5/fXXSUhIYO7cuWi1Wnx9fVmwYAGZmZm8/PLLaLVaNBoN8+bNIzQ01P48CQkJjB07FoDevXvz/vvvX/NzWbduHd27d6d3794sXbr0uhL7smXLePnll+11K6Kiovj+++8JCAhwuG7cuHEUFxfbH0dHRzN9+nSHWHv27IlKpSI8PByLxUJ2drbDl4OWLVtSUFCAVqtFURT7ePPcuXMZMWIECxcurDJe4ajaErtarWbGjBkOx67sSvn9999v2GuVzYgvNNjGzwOuKCW7ITEVd62Gnk0b3LDXq4uSkpIYN24cTz75JEOHDkWv15f7wyPEXzXv3k5Vtq6rw2+//cbo0aPJyspCrVYzbNgwunfvznfffWdPuFeKiooiNTWVjIwMWrVq5XCuonXE8+bNY/To0fTr14+jR48yZcoUVq1axU8//cT06dMdXmPZsmVERETw1ltvkZiYyI4dO/D19QVsc458fX1ZvHixvZJjeno68fHx9O/fn8cff5xNmzaRn5/Pjh07aNOmDZMnT2bv3r3k5eU5JPbCwkJ7rF5eXhQUFFzzM1q5ciUzZsywJ9309HSH56tIRkZGuc/v6qQO8NFHH13zeQoLC/H3v1xPpCzeKxN7ixYtGDJkCB4eHvTv3x9fX19WrVpFYGAgvXr1ksT+F7hESdn8kouoVGp6Nn+QO2PHEtfkLgDS8ov5PS2XXs0a4K6r32Vk1Wo1x44dY8+ePc4ORYgbplu3bixZsoSlS5ei0+mIjIwEoEGDBpw/f77c9WfPnqVhw4aEh4dz4cIFh3PHjh2z74dQ5tSpU9x8880AtG7dutw9V0pKSqJDhw6AbSvjRx55xH7Ozc2N7OxsJk6cyLRp0yguLsZkMjF+/Hiys7N5+OGHWb9+PVqtlgcffJCAgADGjh3L0qVL0Wgc/3ZdOX+pqKjI/uWhIqdOneLEiRPMmTOHJ554ApVKxddff22P6crJj0VFRfZu74iICNLS0hyea9u2bWRmZjocGzduHKNHj7b/d2Vr/epYy17jyi9Qx44d49dff2Xjxo1s2rSJ7Oxs1q1bx7fffsuOHTsYPXo0R48e5aWXXir32qJyLpPYfdwD0WrdaOjfnABPW+u8vpeRPXnyJElJSQA0adKEHTt2MG/ePCdHJcSNFxAQwPz585k6dSoZGRm0b9+eixcv2ieCAWzdupWzZ8/SpUsX7rnnHlauXEl2tq2Xr6ioiGnTppGRkeHwvNHR0ezda5uQe/ToUXvXfUWio6PtPZHJyclMmjTJ4bXT0tJ48803mThxIqWlpSiKwurVq3nggQdYsmQJLVq0YMWKFWzcuJFOnTrx+eefc9ddd7Fo0SKH14mLi2PLli32573WRMyVK1fy/PPP88knn/DJJ5/w+eef8+2332I0GmnTpg0///yzQ4xt27YFYMiQIbz//vuYzWYATp8+zSuvvIJa7ZgyPvroI5YsWWL/7+rEHhcXx7Zt27BaraSmpmK1Wh1a6z4+Pri7u+Pm5oZGoyEwMJD8/HyWLl3Kl19+yZIlS2jdujVz5861T3oUVavzFSZKTUUYzMUEeUdyOvMA4f4xuOls42T1eXz9zJkz9O7dm7Zt27Ju3TrUanWFXZNCuIrmzZszevRoZs6cyeuvv86HH37I7Nmz7d3FYWFhLFy4EI1GQ2RkJC+88AITJkxAo9FQVFTEgw8+SJ8+fRye88UXX+TVV1/l008/xWw2M2vWrEpff8SIEUyZMoWHHnoIi8XClClTOHHiBADt2rXj/fffZ9iwYej1eqKiosjIyKBt27ZMnjwZT09PdDodM2bMQFEUXnjhBd555x3UajUvv/yyw+v84x//4KWXXuIf//gHOp2ON954A4BZs2YxePBgWrduDdjmHKxZs4YffvjBfm94eDitWrXi559/5oknnmDatGkMHjwYvV6Pv7+/fSLtwIEDyczMZOTIkeh0OiwWC/PnzycoKOhP/T+JjY2lc+fODB8+HKvVyrRp0wBYvXo1xcXFDB8+nOHDh9tfp1GjRjzwwAN/6jVEeSpFURRnB1GVsmn/FS13y8g/x9pD79M4MJaz2Yfx0PkwvOsrKIpC5L+/RUHh/GsP1suKcy+++CI9evRg0KBB13W9KyzDqgtc4XOuDcvdruV6l2K5kiVLltC7d2+H0t3VqT5+xtWlot+na+W9qtT5Fnt+iW3cxWgpBcDP09Zdc/hCLhcKShjVqWm9SOpWq5WPPvqInJwcpkyZAiDd7kLUI3379iU8vH4OOwpHdX6MvWwNe4nRNjO0gU8TADYcv1RGNqZ+/EMvLS1l0aJFfP755w4134UQ9YMkdVHGBVrstsRebLIls4jAFgBssI+vhzknsBpgtVpJTk6mcePGeHp68sUXXxASEnLNWbJCCCFcW51vseeVXESj0mI0lwAQ5BVBqcnC1lPptG3oT0Pf2rmt5N+lKArDhg3jrrvuIicnB4A2bdrQoEH9Xq8vhBD1XZ1usSuKQkHJRTzd/CgozUKvcUer0bPlVBqlZotLz4ZXqVT07NkTDw8PLBaLs8MRQghRS9TpxF5szMdsNRHgGUqHRv3sJWXLuuFdbXz93LlzfP7550ydOhWVSsWzzz6LSqWqF5MDhRBCXJ863RVfNr7u7xVKdIOOtAi1LSGKT0xDr1HTq5lrdUtPmzaNt956i/j4eAB73Xch6qNdu3bRvXt3e9WzwYMH8+yzz9qXDmVnZ/PSSy8xevRoRo4cyaRJkxyql+3du5dHH32U0aNHM2TIEJYuXXrdr7106VIGDRrE2rVrr3ndqlWrWLBgwV97g1XYsGGDQxGca1m4cCE9e/bEYDDYj02ePJmtW7c6XNejRw/7z/Hx8fbPdujQoaxfv/4vxblixQoGDx7MsGHD2Lx5c7nzO3bssJ+/ckMfsG1gM2jQoHJximur0y32ssSuUemwWs2o1VoyCkrYfz6bvi3C8NTX6bcHOK4VnT17NgMGDHDY2EKI2mDP6bWcuXjohj5nk+B23Ny08i1RoX7u7gYwc+ZMtm3bZi9GU5XVq1czYMAA1qxZw+DBg6u8ft++fXz22Wd89NFHeHl5kZOTw/Dhw2nevLl9m9zrkZmZyZIlS/j2228xGAyMHDmSHj16OKzXnjdvHgsWLCA6OpqRI0dy/Phx++c6Y8YMabz8BXU685WtYT9wLp795zYwqtt0Np6wlZF1hfH1b7/9lpdeeokffviBNm3aEB4ezrBhw5wdlhC1Un3Z3Q1spVr79evH8uXLq/xcdu3aRaNGjRgxYgQvvPDCdSX2lStX8vDDD9sbFQEBAaxcubLciptXXnmFc+fO2R/7+fnx7rvv2h8fOnSIjh07otfr0ev1NGrUiGPHjjnsyd66dWtyc3MxmUwYDAZ7bfxPPvmEjh07UgdqqNU6dTqxl23XqmBFo9ah07q51Pi6r68vFouF5ORk2rRp4+xwhKjUzU0HVNm6rg71cXc3gAEDBrBr167r+oxWrlzJ0KFDadasGXq9noMHD9K+ffsKry1rHVe0u5ufn1+5669VYrfsfV/5uXp5eVFYWOhwTcuWLRk/fjz+/v60bNmSZs2asXPnTs6ePcuMGTPYt2/fdb1PcVmdTuz5pRfRqPRYFCPebv4oikJ8Yhoh3m60Dy+/xWBtpygKK1euZMCAAXh7e9O/f38OHDhQ4S+UEOJyV3xOTg6PPfaYw+5uv/zyS7nrz549yy233EJGRkaFu7spiuLQvf1nd3cr68aPiYkhJiaGVatWAY67u3l6ejrs7vbhhx/y8MMPExoaSrt27XjwwQf5+OOPGTt2LD4+Pjz//PN/+fPJy8tj69atZGdns2TJEgoLC/nyyy9p3749bm5u9vkIZco2fQkPDyctLc3hy09CQgLBwcEOJWurarFXtbtbfn4+H330EWvWrCE0NJR58+bx6aefcvToUc6fP8/o0aNJSkriyJEjhISEXPfQQ31XZyfPWa0WCkqz0Wps300CvSM4mp7H+bxi+rZoiFpd98Zlli5dyvjx45k9e7b9mCR1IapWn3Z3+zN+/PFHhgwZwqeffsonn3zCihUr2L59O9nZ2bRp04YNGzbYr927d699/Hzw4MF88sknFBcXA5CVlcWUKVMoKSlxeP5Zs2Y57O52ZVIH2+Y3CQkJGAwGCgoKOHXqFDExMfbz7u7ueHp62odAGjRoQH5+Pm+88QbLli1jyZIl9OrVixdeeEGS+p9QZ1vshYYcFMVqX+IW7h/NuuN1bze3svEjlUrF0KFDOXjwIE8//bSToxKi7qkvu7tVZuHChbRq1cph8t/KlSsd9ozw8PDgjjvuYMWKFTz++OMcPXqUQYMG4eXlZX99gI4dOzJs2DAee+wxtFotpaWlTJw4sdzwRVVCQkLsqxIUReH555/Hzc2NnTt3kpCQwIQJE5g8eTKPPfYYbm5u+Pj4MGfOnD/1GqK8Oru7W3L2MTb+8RkqlRpFsfJg58mM+HIf646e5+yrg4n0r/27DqWlpfHcc88xfPjw65rQUt1cYdexusAVPmfZ3a322bhxI56ennTv3r1GXq8+fsbVRXZ3u6RsqVt0SBwKClqNN1tOXaB1qF+dSOpgW6O5bds2fHx8akViF0LUXa1bt5aNYATgAon9pvBbCPQO59eTFyg2Wuhfy7vh09LSMJvNREVF0axZMzZs2CBjR0KIv02SuihTZyfPlSV2Hw/bZJb4xNo/vp6SkkKPHj0YN24cVqttbsBNN90kBRiEEELcMHW3xV6aiUql5qudr9ErZjgbEtPQadT0iQ6t+mYniYiI4O67767z46tCCCFqrzqZ2M0WI0WGPNvEORRU6kASUn6nd7NQvN10zg7PTlEUVq1aRXJyMs899xwqlYr33nvP2WEJIYRwYXUyseeXZgGgKFZAxe5kM4pCrRtfLy0tZebMmWRnZ/Pwww8TEFD3iuYIUZdNmDCh3NpqV7Rr1y6ee+45+zr0oqIiIiMjWbBgAXq9nuzsbObOnUtqaioWi4WGDRsyefJkQkJCANsa9vfeew+z2UxxcTGDBw9m1KhRznxL5YwfPx6ADz/80H7s9ttvZ926dfZHSglfAAAgAElEQVRZ46dOnWL69OksWbIEq9XKwoUL2bp1q71M7dSpU/90fX+r1cr06dM5fvw4er2emTNnOhTpAVv52zVr1qBSqRg/fjz9+/dHURR69+5NkyZNAOjQocN1b9rzd9XNxH5pfB3AXevJxhPpQO0YX1cUhYyMDEJDQ/Hw8OCTTz4hICBAkroQTlAfknqZyjbEufPOO6t1Q5yakJaWZq/Wl5ycXGG54KstWrSInJwcvvzyS9RqNYcOHeLpp59m/fr16HTX37MbHx+P0Whk+fLlHDhwgDlz5vDBBx/Yz+fn57NkyRJ++eUXSkpKuP/+++nfvz/nzp2jTZs2Dl9EakqdT+y+HsFsSEwj0FNPXGSgE6OyJfVx48axfft2duzYgZ+fH3FxcU6NSYiasnJPxYVFYiP60DrctrZ66/HlpOefLndNiE8jbm01EoDEC7s5mLyJoTdPvubrrVq1is2bN1NaWkpmZiZjxoxh48aNnDhxgv/3//4fAwcOpEePHmzfvp2DBw8ya9YsFEUhNDSUBQsW8MQTTxAQEEB+fj4LFy7klVdeITk5GYvFwqOPPsqAAY617wsLC3nllVcoKCggJyeHoUOHctdddzFq1CjWrl2LSqXi3//+t32zmZkzZ/7/9u49Lsf7f+D46+6uu6RyKhZbQ9FsGM1kOR+aKKzSQStjGcaYGGEkImcmfMPXOaeUMIc5Zo4zo2EMbQ3DRqGig7u7+75+f/Tr+nYPFUpqn+fj4aHu+76u63O969H7+lyH9xuAqlWrEh4ezq+//sqcOXMwMjLC29sbExMTvVaxCxYsoFq1akyePJkLFy5gaWnJrVu3iIyMRKlUMnHiRNRqNcbGxoSFhWFt/fSJTMGGOBcuXCiRhjjXrl1jwoQJaDQaTExMmDZtGmFhYXTv3p127dpx5MgRdu/ezYwZM+jYsSP169fHxsaGY8eOsX37dkxNTVm+fDmGhoZ07dr1mfYnNjaWzp07Y2JiwoYNGwgODi70dwMgOjqauLg4DAzy7hFv2rQpsbGxekk9MzNTPhOQz9HRkS+++EL+/syZM7Rt2xbIm3VfuHBB7/OVKlWidu3aZGdnk52dLd8MffHiRe7cuUNAQAAmJiaMGzeO+vXrFznuklDuE7vS8DX+TM2g97tvojQo25v8FQoFDRo04ObNm2RkZIhysIJQyjIzM1m5ciW7du1i9erVbN68mR9//JFVq1bh6uoqf27ixInMnz8fW1tb1q9fT1JSEgA9evTA2dmZdevWyWVpMzIy8PDwoFWrVlSv/r/JwvXr13F1deXDDz+U/2D7+flhb2/P6dOneffddzl16hRff/01fn5+hIeHY2dnR0xMDMuXL8fJyQm1Wk1MTAyQd0p52bJlVKpUiZCQEI4dO4apqSlpaWnExsZy//59PvzwQwBmzpxJQEAA7du354cffmDOnDnMnTtXLxZPa4ize/fuEmmIM3PmTAYOHEi7du3YvXs3ly9ffurP5e+//yYuLk6O6b59+/joo4/YvXs3K1asYPLkyUXuTz6dTsfOnTuJjo7G0NAQV1dXvvzyS0xMTJ66fci7FPrPv8H/PHNauXJloqKiCl1PRkYGZmZm8vdKpZLc3FwMDf+XPq2trXF1dUWr1TJo0CAgr+rewIED6datG6dPn2b06NFs2bKl0G2VlHKd2Guav0lSqhWQUWbX1+/evcv69esZPnw4CoWCoKAgRo0aJR8lCsK/RVEzbIB29j5Ffqbhay1p+FrLYm0zvwaEubk5tra2KBQKqlSpglqt1vvcvXv3sLW1BdC7dlyvXj0g79qsk5MTkNe4xNbWlt9//52FCxcCeTNcDw8P1qxZw759+zAzM5Mbpnh7e7N161ZSUlLo1KkThoaGJCUlMXnyZAA0Go28nfz/AWrUqEFwcDCVK1fmjz/+oFmzZvL/ANWrV5dneImJiSxdupTly5cjSdITTyU/rSFOrVq1uHXr1mOff9aGOFevXqV58+ZAXne5zMxMDhw4IL9fsIhpwcuPXl5ehIaGUr9+ferWrUu1atWKtT/5jh49SmZmpnx9WqfTsWPHDry8vORGNvnX2LOysuSEb2Fh8VhS3r9/Px988IH8WnFm7P9sZKPT6fSS+pEjR0hOTubgwYMABAYG4uDgQOPGjeVr+y1atODOnTtIkvRSHm8ul4k9PfsuFiaWdH/3c3qtOASU3Y1z48aNY8uWLdjZ2eHq6qr3AxcEoXQV949kzZo1uXbtGnXr1mXZsmVygs1fPr/Zi7OzMxkZGSQmJmJra6s3m5s+fTrNmjXDz8+PkydPcvjwYQA++OADZs+ezZ07dwgJCQHyEvjMmTOpXbs2Z86cISUlBUA+4H/48CERERF8//33APTv3x9JkmjQoAHbt28H8jqzXbt2DYD69evz6aef4uDgQFJSEj/99NNT9zV/lty3b1+2bduGg4OD3BCnU6dOgH5DnDfeeIOhQ4fSvXt3qlevLjfEGTp0qF5iz29y4+TkxLfffktKSgoqlUret19//VX+bMGJTd26dZEkieXLl9OnT59n3p/Y2FimTp1Khw4dgLxT41OnTsXLy4u3336bvXv30rt3b3m/mjRpAoC7uzuLFi0iODgYhUJBQkIC06dPZ8+ePfK6izNjd3Bw4NChQ3Tv3p2zZ8/qNbGBvEZdJiYmqFQqFAoF5ubmPHjwgEWLFlG1alU+++wzLl++TO3atV9azZJyl4XUmizUuZlYmb+BRqvj+6TbNLSy4M3qZkUvXEI0Go18hDlp0iQcHBxwcXF5adsXBOHZTJ48mfHjx2NgYICVlRX9+vVj7dq18vve3t5MnDiRPn36oFar+eKLL6hRo4beOjp27EhoaCg7duygatWqKJVKcnJyUKlUdO3alRMnTsh3S4eGhhIcHIxWqwXyuqAV7BxnZmaGg4MD7u7umJqaYmFhQXJyMh4eHhw5cgRfX18sLS0xMTHByMiI4OBgQkNDUavVPHr0iK+//rrQ/S3YECciIqLEGuKEhIQQGRmJiYkJkydPlru+7dixQ777+0l69+7NggULaNWqFcBT9ycoKIjx48fLd+vfu3ePc+fO6d0U+N5776FWq0lISJCb9GzcuBFDQ0PeeOMN+UxJYGAgCxYswMfHB0NDQwwNDYmMjHzm/gbOzs4cP34cX19fJEmSu2+uWrUKGxsbOnfuzIkTJ/D29sbAwAAHBwdat25NkyZNGD16NIcPH0apVDJ9+vRn2u6LKHdNYB7k3GHXuf8ACsxNG+O9XseQ1vYs9CjeqbsXtX//fkaNGkV0dHSFKwVbEZqTlAcVIc6iCUzpSEpK4vLly7i6upKamoqbmxuHDh16JeNcGjGeN28egwcPfuzGvYruX98E5n83zkncSssEKr3U0/BarZaUlBQuXLhQ4RK7IAhly9ramjlz5rBmzRq0Wi1fffXVK5nUS4uvr++/LqmXhnKX2NML3BH/4w1jlAYKOtiVbhnZ/fv34+TkROXKlXFxcSEhIaHQRzMEQRCeh6mpqd4z0v82opFNySh3t24XfNTt20sSH7xphYVJ6R3RbtmyBR8fH6ZNmya/JpK6IAiC8Koqd4k9LTP/sQwlWRplqVebc3V1pXfv3vTt27dUtyMIgiAIJaFcJXZJkniQnVcnPjs371lFZ/uSTexpaWkMHTpUfuTExMSEZcuWPVbAQRAEQRBeReUqsT/SZKJDi5HShJ//tqCKiREtXq9R9ILP4O7du2zdupV169aV6HoFQRAE4WUoVzfPPfz/rm6vVXVgxel03JtYY6h88WOT9PR0MjMzqV27NnZ2dnz77be8++67L7xeQRCEstCpUyesra0xMDBAq9WSlZVFWFgYTZo0QZIkNmzYwM6dO+WCWgMGDJCfW09PT2fmzJlcv35d7gQ3ZcqUJ5aZLSu7d+9m/Pjx7N27l1q18m6eXrhwIZaWlnIRHMirTzBv3jxef/31EutgFx8fz+LFizE0NMTT0xNvb2+99y9dusSkSZNQKpXUrVuXadOmyQV7dDodAwcOpHPnznrjLGnlKrFnqFMBSLqfF6SSuL5+584dOnfuTP369dm2bRsGBga0aNHihdcrCIJQllauXCk//3z06FEWLVrE0qVLiY6OJiEhgdWrV2NsbExqaioDBw6kSpUqNGvWjJEjR+Lr64uzszMAq1evJiQkRK9ITFmLiYnB39+fzZs3M2zYsCI/f+PGjRLpYKfRaJg+fTqxsbFUqlSJPn360LFjR7mgDuR1FBw6dCjt27dn1KhRfP/993LFv2+++Yb09PRn3+FnVK4S+8Ps+wBkZZ8DbPiwBK6v16xZk/fff59GjRqh0+lEjXdBeE6vYne3devWsW/fPnJzczE3N2fhwoXodDrGjRvHX3/9hUajYeLEiVy9epUtW7ag0+kYPnw4KSkprFmzBpVKRd26dZkyZcpj9cyftO6RI0fSt29fWrZsyfnz54mMjCQiIoJJkyZx/fp1dDodI0aMwNHRETc3N+rWrYtKpWLMmDFyJbb8+3y6dOnCoUOHiIiIwMzMjCpVqmBvb8+wYcOYO3cuP/30E5Ik0a9fP7p161ZorP766y8sLCzkca9du1ZO+tWqVeOLL75g48aNWFlZcffuXTmpAwQEBODp6am3Pp1Ox4wZM7h06RIajYZhw4Zhbm7Opk2b5AOA/M56Y8eOJS0tjbS0NOrVq0fLli1xd3cnJSWFQYMGERcX90z7c+PGDdLT0xk0aBDu7u4MHjy4yDas27dvL1YHu/nz55OQkKD32ooVK+RaAklJSdjY2MjNZd577z1Onz6tN95GjRqRlpaGJElkZmbKZ0X27NmDQqF4Ka1wy1Viv5/5FwAKsqlfw4z6NZ7v1NChQ4e4cuUKgwcPRqFQsHLlypdWw1cQhJJTWHe3bt26kZaWxurVqzEwMCAwMJBffvmFX375hTp16jB//nwSExM5ceIEFhYWWFhYEBkZSWpqKiEhIWzduhUzMzPCw8OJjo7G399f3q5Op3viur28vNi6dSstW7Zk69ateHt7ExMTQ7Vq1QgPDyc1NRV/f3927dpFVlYWQ4YM4e233+bEiRP0798fR0dHEhISWLhwIR07dmTq1KlER0djaWkpN0E5fPgwN2/eZNOmTajVary9vWndurWcuPN9+umnqNVqkpOTadu2rdzqNDU1Va9rHeh3estvHpNPqVQ+dhr+4MGDche6lJQU1q1bJzfReZJWrVrRr18/uTmOu7s727dvx8PDo9j7ky82NhZPT0/Mzc1p1qwZ+/fvf6zFbkEKhaLYHeyCgoKeuh7I6/RWcLnKlSuTkZGh95n8A8HIyEjMzc1xdHQkMTGRnTt3EhERweLFiwvdRkkoV4k9/474mw+McW74fIUMcnJyCAoKIjk5GU9PT6ysrERSF4QS8Kp1dzMwMMDIyIiRI0diamrK7du3yc3N5Y8//pBnTQ0bNqRhw4bExcXJjWFu3LiBnZ2d3AHs/fff59ixY3qzudWrVz9x3W3btmX27NmkpaVx+vRpJkyYQFhYGGfOnOH8+fMA5Obmkpqad1kxf5tWVlZERkYSGxuLQqEgNzeX+/fvY2ZmJs8yW7Rowd27d0lMTOTixYsEBATI6ys4I8+Xfyp+3rx53Lx5U659b2ZmRlpaGlWrVpU/e/36daytraldu/Zjnd40Gg179uyhR48e8mtXr16ladOm8tiDgoL48ccf9ZYrWK08fz9tbW3RarXcunWL3bt3s3r1aqKjo4u1P5BX+XPHjh3UqVOH+Ph40tPTWbduHd27d5c7vRWU3+3tSfv1pA52Rc3Y/9npLTMz87EDhGnTprF+/XoaNGjA+vXrmTFjBpUqVeLOnTt88skn3Lp1CyMjI+rUqVNqs/dydd5Zo3sEwO93Kz3z9fUHDx4AebV4lyxZwnfffad3XUQQhPKnsIPyy5cvc+DAAb755hsmTpyITqdDkiS5SxnkJfH8mXD+ZbjXX3+dpKQksrKyADh16hT16tUjKCiIqKgooqKi+O233564bgMDA1xcXAgNDaVLly4olUrq16+Pq6srUVFR/Pe//8XFxUU+lZu/zQULFtCrVy9mz56No6MjkiRRo0YNMjMzuX8/7xLkuXPngLzOaI6OjkRFRbFmzRq6dev22Cy7oBEjRpCcnMyGDRsA8Pf3Z+rUqXISvHfvHosWLcLX15datWpRrVo1vXasa9eu1fs+fwwXL14E8jrVBQYGYmxsLHd6u3Xrlt615II/p969ezN79mzs7OywsLB4pv05fPgwjRs3JioqihUrVhAbG8u9e/e4fPky77zzDvHx8XI73T///JOcnBxq1KiBm5sbMTExcizzO9gVbMwD6P2M8/8VLOlra2vL9evXSUtLIycnh9OnT8utbPNVqVJFPiisWbMmDx48YMyYMcTExBAVFYW7uzv9+vUr1VPy5WrGnu/8bQs6NXit2J8fO3Ysu3fv5tixY1hYWMgdhgRBqLjefPNNKlWqhIeHByqVCisrK5KTk/H19WX8+PH4+/uj1WoZP348v/32m7xc9erVGTZsGH379sXAwAAbGxu++uqrYq0bwNPTky5durB3714gr/75hAkT8Pf3JyMjAz8/v8fu5XFxcWHatGksXboUa2trUlNTMTAwYOLEiXz22WeYm5uj0+l488036dSpE6dOncLPz4+srCy6dOmi13P8nwwMDJg2bRoff/wxXbp0ISAgAK1Wy8cff4yhoSEKhYIhQ4bg4OAAwKxZs5gyZQorV65Eo9FgY2PD1KlT9dbZuXNnjhw5Qp8+fdBqtQwdOpTGjRtjbm6Ol5cXtra2T03O+fuaXzr3afsTFxcHgIeHh7zs5s2b8fLy0ltf7969Wb9+PWFhYSQkJODh4YGZmRmSJDFz5kyAYnewK4qRkRFjx44lMDAQSZLw9PSkVq1a/P7776xbt47Q0FCmTp1KUFAQhoaGGBkZERYW9kzbKAnlqrvb5Ue7yNFlsfzndhwf/vRrKv80Y8YMduzYwdq1a7G1tS3FkZZvFaHrWHlQEeIsuru9HEuXLqV///6oVCq++uor2rRpw0cffVTWwwJKP8aXL1/mwoULcq/1iqyku7uVq1PxAPezDXFuWKfQz2RkZLBq1Sr5Gs+oUaOIj48XSV0QhHKlcuXKeHt7y73AC7tJrKKpWrXqY3fjC8VT7k7FLzxpwzr/wq+vBwcHs3HjRqpXr06vXr2KfBRCEAThVeTv7693N/6/yWuvFf9yq6Cv3CX2LE1lHN+0fOz1gs+gBwcHU7t2bbp27fqyhycIgiAIz0SSpBJ9OqtcnYrP1UIH2xoY/aOM7MmTJ2nbti1XrlwBwMbGhq+//hoTE5OyGKYgVHgGBgby3ceCILwYrVZbosXRytWMXWkAHexsHnv97t27XLlyhRMnTmBvb18GIxOEfxdDQ0Oys7PJyspCqVS+crUgNBrNY880CyVLxPjFSZKEVqtFq9XKFepKQrmasesk6GKf9wjFTz/9RHZ2NgBubm6cPHmS/v37l+XwBOFfxdzcHJVK9coldcgr/SmULhHjF6dQKFCpVCXeYKdczdgf5RrSwNKc7777Dn9/fz7//HP5+Uo7O7syHp0g/PuU5CyjpL2qj+JVJCLGr6ZSm7HrdDpCQkLw8fEhICCA69ev672/efNmPDw88Pb25tChQ8Vbp2SOQqGgffv2ODs74+bmVhpDFwRBEIRyq9QOtw8cOEBOTg7R0dGcPXuWGTNmyJWGUlJSiIqKYsuWLajVavz8/GjdunWRR3+3r2RAJzA1NWXTpk2lNXRBEARBKLdKLbGfOXOGtm3bAtCsWTMuXLggv3f+/HmaN2+OSqVCpVJhY2PD5cuX5aYC/5RfaObGmSQePXr0Sl7TqyjUanVZD+FfQcS59IkYlz4R49KTf2Pi8xSHLbXEnpGRoVe/WKlUkpubi6GhYbFa3xWk0WgAGPL553LjAaF0FDwAE0qPiHPpEzEufSLGpU+j0Tzzo9ulltj/2d5Op9PJN9oUp/VdQZUrV6Zhw4YYGRmJ2bogCIJQ4UmShEajea56/KWW2B0cHDh06BDdu3fn7NmzNGzYUH6vadOmfPPNN6jVanJyckhKStJ7/58MDAxK/HEAQRAEQXiVPW+RtVLr7qbT6QgNDSUxMRFJkggPD+fIkSPY2NjQuXNnNm/eTHR0NJIkMWjQIFH+VRAEQRBKQLlo2yoIgiAIQvGUq8pzgiAIgiAUTiR2QRAEQahAXrnEXhoV6wR9RcV49erVeHl54eXlxaJFi8polOVbUTHO/8yAAQPYuHFjGYyw/CsqxocPH8bb2xtvb29CQ0Of63lgoeg4r1ixAg8PDzw9Pdm/f38ZjbJiOHfuHAEBAY+9Hh8fj6enJz4+PmzevLnoFUmvmL1790rBwcGSJEnSzz//LA0ePFh+Lzk5WXJzc5PUarX04MED+Wvh2RQW4z///FNyd3eXcnNzJa1WK/n4+EiXLl0qq6GWW4XFON/cuXOl3r17Sxs2bHjZw6sQCovxw4cPJVdXV+nevXuSJEnSsmXL5K+FZ1NYnNPT06X27dtLarVaSktLkzp06FBWwyz3li1bJrm5uUleXl56r+fk5EhdunSR0tLSJLVaLXl4eEjJycmFruuVm7EXt2Kdubm5XLFOeDaFxfi1115j+fLlKJVKuee2sbFxWQ213CosxgB79uxBoVDQrl27shhehVBYjH/++WcaNmzIzJkz8fPzw9LSkurVq5fVUMu1wuJcqVIlateuTXZ2NtnZ2aLOyAuwsbFh4cKFj72elJSEjY0NVapUQaVS8d5773H69OlC1/XKtWYqyYp1wpMVFmMjIyOqV6+OJEnMmjWLt99+m3r16pXhaMunwmKcmJjIzp07iYiIYPHixWU4yvKtsBinpqby448/sm3bNkxNTfn4449p1qyZ+F1+DoXFGcDa2hpXV1e0Wi2DBg0qq2GWe127duXmzZuPvf48ee+VS+wlWbFOeLLCYgx59Z/Hjx9P5cqVmTRpUlkMsdwrLMbbtm3jzp07fPLJJ9y6dQsjIyPq1KkjZu/PqLAYV61alSZNmmBlZQVAixYtuHTpkkjsz6GwOB85coTk5GQOHjwIQGBgIA4ODk/t+yE8u+fJe6/cqXgHBweOHDkC8MSKdWfOnEGtVvPw4cMiK9YJT1ZYjCVJYsiQIdjb2zNlyhSUSmVZDbNcKyzGY8aMISYmhqioKNzd3enXr59I6s+hsBg3btyYxMRE7t+/T25uLufOncPOzq6shlquFRbnKlWqYGJigkqlwtjYGHNzcx48eFBWQ62QbG1tuX79OmlpaeTk5HD69GmaN29e6DKv3Izd2dmZ48eP4+vrK1esW7VqlVyxLiAgAD8/PyRJIigoSFz/fQ6FxVin03Hq1ClycnI4evQoACNHjizyF0nQV9TvsfDiiorxqFGjGDBgAAAuLi5iEvCciorziRMn8Pb2xsDAAAcHB1q3bl3WQ64QduzYQVZWFj4+PowdO5bAwEAkScLT05NatWoVuqyoPCcIgiAIFcgrdypeEARBEITnJxK7IAiCIFQgIrELgiAIQgUiErsgCIIgVCAisQuCIAhCBfLKPe4mCBXRzZs3cXFxwdbWVu/1JUuWYG1t/cRl8stLDhs27Lm3GxcXx4wZM+RtPHr0iJYtWzJp0iS9okTFsWDBAho3biw/dhoVFQVAr1692L59+3OPESAgIIDbt29jamoK5FXbeuONN5gzZw6WlpZPXW7z5s2Ympri5ub2QtsXhIpEJHZBeElq1qz5wgnweXTq1IkZM2YAoNVq8fX1JTY2Fl9f32daz5dffil/ferUKfnrktqnqVOn4ujoCORVNxs+fDirVq1i9OjRT10mISGBli1blsj2BaGiEIldEMpYYmIiYWFhZGVlcf/+fQYOHEifPn3k9zUaDePHj+e3334DwM/PD29vb+7evUtISAi3b99GoVAwatQonJycCt2WUqmkRYsW8rq2bNnCqlWrUCgUvPPOO0ycOBGVSvXE7Y0dO5aWLVvy66+/AuDl5UVMTAz29vZcvHiRDh06sG3bNiwtLUlLS8PNzY1Dhw7xww8/EBERQW5uLq+//jphYWFUq1at0HFmZWWRmpoqlyb97rvvWLVqFY8ePSInJ4fw8HAePXpEfHw8J0+exMrKikaNGj1zPAShIhLX2AXhJUlOTqZXr17yv+XLlwMQExPDkCFD2LJlC2vXrmXWrFl6y/3888+kp6ezbds2li5dKnd2mjZtGp6ensTFxREZGUlISEiRzSFSU1M5duwYzZo148qVKyxZsoSoqCh27NhBpUqVWLRo0VO3l2/ChAnyuPMZGhri4uLCnj17ANi3bx/Ozs48fPiQuXPnsmLFCrZt20abNm2YM2fOE8c2YcIEevbsSZs2bfDx8cHJyYl+/fqh0+nYtGkTS5Ys4dtvv2XAgAEsW7YMJycnOnXqxPDhw2nbtu1zxUMQKiIxYxeEl+Rpp+LHjh3L0aNHWbp0KYmJiWRlZem936BBA65evUpgYCDt2rVjzJgxAJw4cYI//viDiIgIAHJzc7lx4waNGjXSWz4+Pp5evXohSRKSJOHs7Iybmxvr16+nY8eO8uzZx8eHcePGMXDgwCduryg9e/Zk+vTp+Pv7s3PnToKCgjh37hx///03ffv2BfJOsVepUuWJy+efik9ISGD48OE4OzujUqkAWLx4MfHx8Vy9epVTp05hYPD4nKS48RCEik4kdkEoYyNGjMDCwoKOHTvSvXt3du7cqfd+tWrV2LVrF8ePH+fw4cO4u7uza9cudDoda9asoWrVqkDeGYEaNWo8tv6C19gL0ul0et9LkkRubu5Tt1eUpk2bkp6ezvnz57lz5w7NmzfnwIEDODg4sGTJEiCvc2DBTlVP4uDgQEBAAKNGjWLr1q2o1Wp69+5Nz549ef/997G3t2f9+vVP3J/ixEMQKjpxKl4Qytjx48cZPrlWDJIAAAGvSURBVHw4Xbp0kbtoabVa+f2DBw8yevRoOnTowIQJEzA1NeXvv/+mVatWbNiwAYDff/+dHj16kJ2dXezttmzZkvj4eNLS0oC8O8wdHR2fur2C8nty/1OPHj2YNGkSrq6uALz77rucPXuWq1evAvCf//znsUsNT9K/f38yMzOJjo7m2rVrKBQKBg8ejKOjI/v375fjo1Qq5a9fNB6CUFGIGbsglLFhw4bh5+eHsbExb731FnXq1OHmzZvy++3atWPfvn24urpibGxMz549sbe3Z8KECYSEhNCjRw8AZs2ahZmZWbG3+9ZbbzFo0CACAgLQaDS88847TJ48GWNj4ydur6DOnTvTq1cv4uLi9F7v2bMnCxYsYP78+QBYWVkRHh7OiBEj0Ol01KpVi9mzZxc5NpVKxYgRIwgPD2f//v00atSIbt26oVAoaNOmDWfOnAHAycmJefPmYW5u/sLxEISKQnR3EwRBEIQKRJyKFwRBEIQKRCR2QRAEQahARGIXBEEQhApEJHZBEARBqEBEYhcEQRCECkQkdkEQBEGoQERiFwRBEIQKRCR2QRAEQahA/g8iBF7crLiswwAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf)" ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf, plot = 'confusion_matrix')" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf, plot = 'boundary')" ] }, { "cell_type": "code", "execution_count": 42, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf, plot = 'feature')" ] }, { "cell_type": "code", "execution_count": 43, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf, plot = 'pr')" ] }, { "cell_type": "code", "execution_count": 44, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_model(rf, plot = 'class_report')" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "eca7ccd44fe3487c9369e4a8b65f1374", "version_major": 2, "version_minor": 0 }, "text/plain": [ "interactive(children=(ToggleButtons(description='Plot Type:', icons=('',), options=(('Hyperparameters', 'param…" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "evaluate_model(rf)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 10. Interpret Model" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [], "source": [ "catboost = create_model('catboost', cross_validation=False)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "interpret_model(catboost)" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "interpret_model(catboost, plot = 'correlation')" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "
\n", "
\n", " Visualization omitted, Javascript library not loaded!
\n", " Have you run `initjs()` in this notebook? If this notebook was from another\n", " user you must also trust this notebook (File -> Trust notebook). If you are viewing\n", " this notebook on github the Javascript has been stripped for security. If you are using\n", " JupyterLab this error is because a JupyterLab extension has not yet been written.\n", "
\n", " " ], "text/plain": [ "" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "interpret_model(catboost, plot = 'reason', observation = 12)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 11. AutoML()" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "LogisticRegression(C=5.5600000000000005, class_weight='balanced', dual=False,\n", " fit_intercept=True, intercept_scaling=1, l1_ratio=None,\n", " max_iter=100, multi_class='auto', n_jobs=-1, penalty='l2',\n", " random_state=123, solver='lbfgs', tol=0.0001, verbose=0,\n", " warm_start=False)" ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" } ], "source": [ "best = automl(optimize = 'Recall')\n", "best" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 12. Predict Model" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
ModelAccuracyAUCRecallPrec.F1KappaMCC
0Logistic Regression0.84160.90480.7680.81360.79010.66310.6638
\n", "
" ], "text/plain": [ " Model Accuracy AUC Recall Prec. F1 Kappa \\\n", "0 Logistic Regression 0.8416 0.9048 0.768 0.8136 0.7901 0.6631 \n", "\n", " MCC \n", "0 0.6638 " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
WeekofPurchasePriceCHPriceMMDiscCHDiscMMLoyalCHSalePriceMMSalePriceCHPriceDiffPctDiscMM...Store7_NoStore7_YesSTORE_0STORE_1STORE_2STORE_3STORE_4PurchaseLabelScore
0260.01.862.180.00.700.9593051.481.86-0.380.321101...1.00.00.00.01.00.00.0000.1873
1229.01.691.690.00.000.7952001.691.690.000.000000...1.00.00.00.01.00.00.0000.1914
2261.01.862.130.00.240.5889651.891.860.030.112676...0.01.01.00.00.00.00.0000.2532
3247.01.992.230.00.000.0036892.231.990.240.000000...1.00.00.00.00.01.00.0110.9127
4271.01.992.090.10.400.9736121.691.89-0.200.191388...1.00.00.00.00.01.00.0000.1895
\n", "

5 rows × 31 columns

\n", "
" ], "text/plain": [ " WeekofPurchase PriceCH PriceMM DiscCH DiscMM LoyalCH SalePriceMM \\\n", "0 260.0 1.86 2.18 0.0 0.70 0.959305 1.48 \n", "1 229.0 1.69 1.69 0.0 0.00 0.795200 1.69 \n", "2 261.0 1.86 2.13 0.0 0.24 0.588965 1.89 \n", "3 247.0 1.99 2.23 0.0 0.00 0.003689 2.23 \n", "4 271.0 1.99 2.09 0.1 0.40 0.973612 1.69 \n", "\n", " SalePriceCH PriceDiff PctDiscMM ... Store7_No Store7_Yes STORE_0 \\\n", "0 1.86 -0.38 0.321101 ... 1.0 0.0 0.0 \n", "1 1.69 0.00 0.000000 ... 1.0 0.0 0.0 \n", "2 1.86 0.03 0.112676 ... 0.0 1.0 1.0 \n", "3 1.99 0.24 0.000000 ... 1.0 0.0 0.0 \n", "4 1.89 -0.20 0.191388 ... 1.0 0.0 0.0 \n", "\n", " STORE_1 STORE_2 STORE_3 STORE_4 Purchase Label Score \n", "0 0.0 1.0 0.0 0.0 0 0 0.1873 \n", "1 0.0 1.0 0.0 0.0 0 0 0.1914 \n", "2 0.0 0.0 0.0 0.0 0 0 0.2532 \n", "3 0.0 0.0 1.0 0.0 1 1 0.9127 \n", "4 0.0 0.0 1.0 0.0 0 0 0.1895 \n", "\n", "[5 rows x 31 columns]" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pred_holdouts = predict_model(lr)\n", "pred_holdouts.head()" ] }, { "cell_type": "code", "execution_count": 30, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
IdWeekofPurchaseStoreIDPriceCHPriceMMDiscCHDiscMMSpecialCHSpecialMMLoyalCHSalePriceMMSalePriceCHPriceDiffStore7PctDiscMMPctDiscCHListPriceDiffSTORELabelScore
0123711.751.990.000.0000.5000001.991.750.24No0.0000000.0000000.24100.4742
1223911.751.990.000.3010.6000001.691.75-0.06No0.1507540.0000000.24110.5433
2324511.862.090.170.0000.6800002.091.690.40No0.0000000.0913980.23100.1670
3422711.691.690.000.0000.4000001.691.690.00No0.0000000.0000000.00110.7475
4522871.691.690.000.0000.9565351.691.690.00Yes0.0000000.0000000.00000.0492
\n", "
" ], "text/plain": [ " Id WeekofPurchase StoreID PriceCH PriceMM DiscCH DiscMM SpecialCH \\\n", "0 1 237 1 1.75 1.99 0.00 0.0 0 \n", "1 2 239 1 1.75 1.99 0.00 0.3 0 \n", "2 3 245 1 1.86 2.09 0.17 0.0 0 \n", "3 4 227 1 1.69 1.69 0.00 0.0 0 \n", "4 5 228 7 1.69 1.69 0.00 0.0 0 \n", "\n", " SpecialMM LoyalCH SalePriceMM SalePriceCH PriceDiff Store7 PctDiscMM \\\n", "0 0 0.500000 1.99 1.75 0.24 No 0.000000 \n", "1 1 0.600000 1.69 1.75 -0.06 No 0.150754 \n", "2 0 0.680000 2.09 1.69 0.40 No 0.000000 \n", "3 0 0.400000 1.69 1.69 0.00 No 0.000000 \n", "4 0 0.956535 1.69 1.69 0.00 Yes 0.000000 \n", "\n", " PctDiscCH ListPriceDiff STORE Label Score \n", "0 0.000000 0.24 1 0 0.4742 \n", "1 0.000000 0.24 1 1 0.5433 \n", "2 0.091398 0.23 1 0 0.1670 \n", "3 0.000000 0.00 1 1 0.7475 \n", "4 0.000000 0.00 0 0 0.0492 " ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "new_data = data.copy()\n", "new_data.drop(['Purchase'], axis=1, inplace=True)\n", "predict_new = predict_model(best, data=new_data)\n", "predict_new.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 13. Save / Load Model" ] }, { "cell_type": "code", "execution_count": 31, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Transformation Pipeline and Model Succesfully Saved\n" ] } ], "source": [ "save_model(best, model_name='best-model')" ] }, { "cell_type": "code", "execution_count": 32, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Transformation Pipeline and Model Sucessfully Loaded\n", "[Pipeline(memory=None,\n", " steps=[('dtypes',\n", " DataTypes_Auto_infer(categorical_features=[],\n", " display_types=True, features_todrop=[],\n", " ml_usecase='classification',\n", " numerical_features=[], target='Purchase',\n", " time_features=[])),\n", " ('imputer',\n", " Simple_Imputer(categorical_strategy='not_available',\n", " numeric_strategy='mean',\n", " target_variable=None)),\n", " ('new_levels1',\n", " New_Catagorical_L...\n", " ('group', Empty()), ('nonliner', Empty()), ('scaling', Empty()),\n", " ('P_transform', Empty()), ('pt_target', Empty()),\n", " ('binn', Empty()), ('rem_outliers', Empty()),\n", " ('cluster_all', Empty()), ('dummy', Dummify(target='Purchase')),\n", " ('fix_perfect', Empty()), ('clean_names', Clean_Colum_Names()),\n", " ('feature_select', Empty()), ('fix_multi', Empty()),\n", " ('dfs', Empty()), ('pca', Empty())],\n", " verbose=False), LogisticRegression(C=5.5600000000000005, class_weight='balanced', dual=False,\n", " fit_intercept=True, intercept_scaling=1, l1_ratio=None,\n", " max_iter=100, multi_class='auto', n_jobs=-1, penalty='l2',\n", " random_state=123, solver='lbfgs', tol=0.0001, verbose=0,\n", " warm_start=False)]\n" ] } ], "source": [ "loaded_bestmodel = load_model('best-model')\n", "print(loaded_bestmodel)" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
Pipeline(memory=None,\n",
       "         steps=[('dtypes',\n",
       "                 DataTypes_Auto_infer(categorical_features=[],\n",
       "                                      display_types=True, features_todrop=[],\n",
       "                                      ml_usecase='classification',\n",
       "                                      numerical_features=[], target='Purchase',\n",
       "                                      time_features=[])),\n",
       "                ('imputer',\n",
       "                 Simple_Imputer(categorical_strategy='not_available',\n",
       "                                numeric_strategy='mean',\n",
       "                                target_variable=None)),\n",
       "                ('new_levels1',\n",
       "                 New_Catagorical_L...\n",
       "                ('group', Empty()), ('nonliner', Empty()), ('scaling', Empty()),\n",
       "                ('P_transform', Empty()), ('pt_target', Empty()),\n",
       "                ('binn', Empty()), ('rem_outliers', Empty()),\n",
       "                ('cluster_all', Empty()), ('dummy', Dummify(target='Purchase')),\n",
       "                ('fix_perfect', Empty()), ('clean_names', Clean_Colum_Names()),\n",
       "                ('feature_select', Empty()), ('fix_multi', Empty()),\n",
       "                ('dfs', Empty()), ('pca', Empty())],\n",
       "         verbose=False)
DataTypes_Auto_infer(ml_usecase='classification', target='Purchase')
Simple_Imputer(categorical_strategy='not_available', numeric_strategy='mean',\n",
       "               target_variable=None)
New_Catagorical_Levels_in_TestData(replacement_strategy='least frequent',\n",
       "                                   target='Purchase')
Empty()
Empty()
Empty()
Empty()
New_Catagorical_Levels_in_TestData(replacement_strategy='least frequent',\n",
       "                                   target='Purchase')
Make_Time_Features(list_of_features=None)
Empty()
Empty()
Empty()
Empty()
Empty()
Empty()
Empty()
Empty()
Dummify(target='Purchase')
Empty()
Clean_Colum_Names()
Empty()
Empty()
Empty()
Empty()
" ], "text/plain": [ "Pipeline(memory=None,\n", " steps=[('dtypes',\n", " DataTypes_Auto_infer(categorical_features=[],\n", " display_types=True, features_todrop=[],\n", " ml_usecase='classification',\n", " numerical_features=[], target='Purchase',\n", " time_features=[])),\n", " ('imputer',\n", " Simple_Imputer(categorical_strategy='not_available',\n", " numeric_strategy='mean',\n", " target_variable=None)),\n", " ('new_levels1',\n", " New_Catagorical_L...\n", " ('group', Empty()), ('nonliner', Empty()), ('scaling', Empty()),\n", " ('P_transform', Empty()), ('pt_target', Empty()),\n", " ('binn', Empty()), ('rem_outliers', Empty()),\n", " ('cluster_all', Empty()), ('dummy', Dummify(target='Purchase')),\n", " ('fix_perfect', Empty()), ('clean_names', Clean_Colum_Names()),\n", " ('feature_select', Empty()), ('fix_multi', Empty()),\n", " ('dfs', Empty()), ('pca', Empty())],\n", " verbose=False)" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn import set_config\n", "set_config(display='diagram')\n", "loaded_bestmodel[0]" ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [], "source": [ "from sklearn import set_config\n", "set_config(display='text')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 14. Deploy Model" ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model Succesfully Deployed on AWS S3\n" ] } ], "source": [ "deploy_model(best, model_name = 'best-aws', authentication = {'bucket' : 'pycaret-test'})" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 15. Get Config / Set Config" ] }, { "cell_type": "code", "execution_count": 36, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
WeekofPurchasePriceCHPriceMMDiscCHDiscMMLoyalCHSalePriceMMSalePriceCHPriceDiffPctDiscMM...SpecialCH_1SpecialMM_0SpecialMM_1Store7_NoStore7_YesSTORE_0STORE_1STORE_2STORE_3STORE_4
584264.01.862.130.370.00.8361602.131.490.640.000000...1.01.00.00.01.01.00.00.00.00.0
751232.01.792.090.000.00.4000002.091.790.300.000000...0.01.00.01.00.00.00.00.00.01.0
462228.01.691.690.000.00.5840001.691.690.000.000000...0.01.00.00.01.01.00.00.00.00.0
7234.01.751.990.000.40.9777461.591.75-0.160.201005...1.01.00.00.01.01.00.00.00.00.0
161269.01.992.090.100.00.9780102.091.890.200.000000...0.01.00.01.00.00.00.00.00.01.0
\n", "

5 rows × 28 columns

\n", "
" ], "text/plain": [ " WeekofPurchase PriceCH PriceMM DiscCH DiscMM LoyalCH SalePriceMM \\\n", "584 264.0 1.86 2.13 0.37 0.0 0.836160 2.13 \n", "751 232.0 1.79 2.09 0.00 0.0 0.400000 2.09 \n", "462 228.0 1.69 1.69 0.00 0.0 0.584000 1.69 \n", "7 234.0 1.75 1.99 0.00 0.4 0.977746 1.59 \n", "161 269.0 1.99 2.09 0.10 0.0 0.978010 2.09 \n", "\n", " SalePriceCH PriceDiff PctDiscMM ... SpecialCH_1 SpecialMM_0 \\\n", "584 1.49 0.64 0.000000 ... 1.0 1.0 \n", "751 1.79 0.30 0.000000 ... 0.0 1.0 \n", "462 1.69 0.00 0.000000 ... 0.0 1.0 \n", "7 1.75 -0.16 0.201005 ... 1.0 1.0 \n", "161 1.89 0.20 0.000000 ... 0.0 1.0 \n", "\n", " SpecialMM_1 Store7_No Store7_Yes STORE_0 STORE_1 STORE_2 STORE_3 \\\n", "584 0.0 0.0 1.0 1.0 0.0 0.0 0.0 \n", "751 0.0 1.0 0.0 0.0 0.0 0.0 0.0 \n", "462 0.0 0.0 1.0 1.0 0.0 0.0 0.0 \n", "7 0.0 0.0 1.0 1.0 0.0 0.0 0.0 \n", "161 0.0 1.0 0.0 0.0 0.0 0.0 0.0 \n", "\n", " STORE_4 \n", "584 0.0 \n", "751 1.0 \n", "462 0.0 \n", "7 0.0 \n", "161 1.0 \n", "\n", "[5 rows x 28 columns]" ] }, "execution_count": 36, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train = get_config('X_train')\n", "X_train.head()" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "123" ] }, "execution_count": 37, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_config('seed')" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [], "source": [ "from pycaret.classification import set_config\n", "set_config('seed', 999)" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "999" ] }, "execution_count": 39, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_config('seed')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 16. MLFlow UI" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "# !mlflow ui" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# End\n", "Thank you. For more information / tutorials on PyCaret, please visit https://www.pycaret.org" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.4" } }, "nbformat": 4, "nbformat_minor": 2 }