{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Abstract" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The aim of this notebook is to illustrate how analysts and data scientists can take advantage of [Vortexa's Python SDK](https://github.com/VorTECHsa/python-sdk) to gain real-time market insights of the oil industry and use Vortexa's data to build powerful predictive models. We will focus on United States Crude oil exports, which have skyrocketed the last couple of years and constitute on of the main drivers of the whole oil industry. In the first part of the notebook, we will familiarize with using the SDK and do an initial **exploratory analysis** of our data to get a sense of their nature. In the second part, we will build a **forecasting model** to predict US Crude exports for the next couple of months. Let's start!\n", "\n", "PS: Please note that the results on this notebook represent Vortexa's data as of 24th February 2020. These data are constantly evolving, refined and improved, therefore some (small) deviations in the numbers presented should be expected in future runs of the notebook. " ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np\n", "from datetime import datetime\n", "from dateutil.relativedelta import relativedelta\n", "import matplotlib.pyplot as plt\n", "from vortexasdk import CargoMovements, VesselMovements, Products, Geographies\n", "\n", "import warnings\n", "warnings.filterwarnings(\"ignore\")\n", "\n", "pd.options.display.max_columns = None\n", "pd.options.display.max_colwidth = 200\n", "pd.options.display.max_rows = 200" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Exploratory Analysis" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Fetching Data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first step is to define the limits of the time period we are interested to look at. Note, that the SDK can return a maximum of 4 years of historical data. In addition, since every entity that lives in Vortexa API is associated with a unique ID, we will need to retrieve the relevant IDs for our analysis. " ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Start date: 2016-02-01 00:00:00\n", "End date: 2020-02-01 00:00:00\n" ] } ], "source": [ "# Define filter limits\n", "END_DATE = datetime(2020,2,1)\n", "START_DATE = datetime(2016,2,1)\n", "print('Start date: {}'.format(START_DATE))\n", "print('End date: {}'.format(END_DATE))" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2020-02-26 10:50:24,794 vortexasdk.operations — INFO — Searching Geographies with params: {'term': ['united states']}\n", "2020-02-26 10:50:24,796 vortexasdk.client — INFO — Creating new VortexaClient\n", "2020-02-26 10:50:25,593 vortexasdk.client — INFO — 1 Results to retreive. Sending 1 post requests in parallel using 6 threads.\n", "United States polygon id: 2d92cc08f22524dba59f6a7e340f132a9da0ce9573cca968eb8e3752ef17a963\n" ] } ], "source": [ "# Find United States ID\n", "us = [g.id for g in Geographies().search('united states').to_list() if 'country' in g.layer]\n", "print('United States polygon id: {}'.format(us[0]))" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "# Check we've only got one ID for US\n", "assert len(us) == 1" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2020-02-26 10:50:26,151 vortexasdk.operations — INFO — Searching Products with params: {'term': ['crude'], 'ids': [], 'product_parent': [], 'allowTopLevelProducts': True}\n", "2020-02-26 10:50:26,947 vortexasdk.client — INFO — 8 Results to retreive. Sending 1 post requests in parallel using 6 threads.\n" ] }, { "data": { "text/plain": [ "['6f11b0724c9a4e85ffa7f1445bc768f054af755a090118dcf99f14745c261653']" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Find Crude ID\n", "crude = [p.id for p in Products().search('crude').to_list() if p.name=='Crude']\n", "crude" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "# Check we've only got one Crude ID\n", "assert len(crude) == 1" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "# Define the set of columns you want to be returned. For the complete set of available columns check the link:\n", "\n", "# https://vortechsa.github.io/python-sdk/endpoints/cargo_movements/#notes\n", "\n", "# Note that not all of the columns defined below are directly necessary for our analysis, but is a good opportunity \n", "# for beginner users of the SDK to get a grasp of the highly non-flat structure of the data and the level of detail\n", "# the data can dig into.\n", "\n", "required_columns = [\n", " # A cargo movement can be carried by multiple vessels across various STS transfers. You can find all the vessels that\n", " # the cargo was onboard by inspecting the 'vessels.0', 'vessels.1' columns etc.\n", " # The 'vessels.0' columns shows the primary vessel associated with the cargo movement\n", " 'cargo_movement_id',\n", " 'vessels.0.name',\n", " 'vessels.0.imo',\n", " 'vessels.0.vessel_class',\n", " 'vessels.0.dwt',\n", " 'vessels.1.name',\n", " 'vessels.1.imo',\n", " 'vessels.1.vessel_class',\n", " 'vessels.1.dwt',\n", " 'vessels.2.name',\n", " 'vessels.2.imo',\n", " 'vessels.2.vessel_class',\n", " 'vessels.2.dwt',\n", " # Bring the loading Geography and Timestamps\n", " 'events.cargo_port_load_event.0.location.terminal.label',\n", " 'events.cargo_port_load_event.0.location.port.label',\n", " 'events.cargo_port_load_event.0.location.country.label',\n", " 'events.cargo_port_load_event.0.location.trading_region.label',\n", " 'events.cargo_port_load_event.0.location.shipping_region.label',\n", " 'events.cargo_port_load_event.1.location.port.label',\n", " 'events.cargo_port_load_event.2.location.port.label',\n", " 'events.cargo_port_load_event.0.start_timestamp',\n", " 'events.cargo_port_load_event.0.end_timestamp',\n", " 'events.cargo_port_load_event.1.end_timestamp',\n", " 'events.cargo_port_load_event.2.end_timestamp',\n", " # Bring STS associated info\n", " 'events.cargo_sts_event.0.start_timestamp',\n", " 'events.cargo_sts_event.0.location.sts_zone.label',\n", " 'events.cargo_sts_event.1.start_timestamp',\n", " 'events.cargo_sts_event.1.location.sts_zone.label',\n", " 'events.cargo_sts_event.2.start_timestamp',\n", " 'events.cargo_sts_event.2.location.sts_zone.label',\n", " # Bring the discharge Geography and Timestamps\n", " 'events.cargo_port_unload_event.0.location.terminal.label',\n", " 'events.cargo_port_unload_event.0.location.port.label',\n", " 'events.cargo_port_unload_event.0.location.country.label',\n", " 'events.cargo_port_unload_event.0.location.region.label',\n", " 'events.cargo_port_unload_event.0.location.trading_subregion.label',\n", " 'events.cargo_port_unload_event.0.location.shipping_region.label',\n", " 'events.cargo_port_unload_event.0.location.trading_region.label', \n", " 'events.cargo_port_unload_event.0.start_timestamp',\n", " 'events.cargo_port_unload_event.0.end_timestamp',\n", " 'events.cargo_port_unload_event.1.location.port.label',\n", " 'events.cargo_port_unload_event.1.end_timestamp',\n", " 'events.cargo_port_unload_event.2.location.port.label',\n", " 'events.cargo_port_unload_event.2.end_timestamp',\n", " #Bring any corporate information associated with the primary vessel\n", " 'vessels.0.corporate_entities.charterer.label',\n", " # Bring product information and quantity\n", " 'product.grade.label',\n", " 'product.category.label',\n", " 'product.group_product.label',\n", " 'product.group.label',\n", " 'quantity',\n", " # Is the vessel in transit, has it already discharged, or is it in floating storage?\n", " 'status'\n", " ]\n", "\n", "\n", "# Define a function to perform some basic manipulations in the data for ease of use\n", "def prepare_cms(cms, ignore_intra = True):\n", " \"\"\" \n", " Performs some basic data manipulation to the cargo movements DataFrame. By default, intra cargo movements,\n", " i.e. cargo movements that start and end in the same country are filtered out.\n", " \"\"\" \n", " # Convert date column to pandas datetime type\n", " cols = [c for c in cms.columns if 'timestamp' in c]\n", " for c in cols:\n", " cms[c] = pd.to_datetime(cms[c]).dt.tz_localize(None)\n", " \n", " # Rename columns for ease of use\n", " cms = cms.rename(columns = {'events.cargo_port_load_event.0.start_timestamp': 'loading_timestamp',\n", " 'events.cargo_port_load_event.0.end_timestamp': 'start_timestamp',\n", " 'events.cargo_port_unload_event.0.start_timestamp': 'unloading_timestamp',\n", " 'events.cargo_port_unload_event.0.end_timestamp': 'end_timestamp',\n", " 'events.cargo_port_load_event.0.location.port.label': 'loading_port',\n", " 'events.cargo_port_load_event.0.location.trading_region.label': 'loading_trading_region',\n", " 'events.cargo_port_load_event.0.location.country.label': 'loading_country',\n", " 'events.cargo_port_unload_event.0.location.country.label': 'unloading_country',\n", " 'product.category.label': 'product_category'\n", " })\n", " \n", " # Calculate loading week and month\n", " cms['loading_week'] = cms['start_timestamp'].map(lambda x: x.to_period('W').start_time.date() if not pd.isnull(x) else x)\n", " cms['loading_month'] = cms['start_timestamp'].map(lambda x: x.to_period('M').start_time.date() if not pd.isnull(x) else x)\n", " \n", " # Depending on user input keep or ignore intra cargo movements, i.e. movements that start and end in the \n", " # same country\n", " if ignore_intra:\n", " cms = cms.loc[cms['loading_country'] != cms['unloading_country']]\n", " \n", " return cms" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2020-02-26 10:50:27,542 vortexasdk.operations — INFO — Searching CargoMovements with params: {'filter_activity': 'loading_end', 'filter_time_min': '2016-02-01T00:00:00.000Z', 'filter_time_max': '2020-02-01T00:00:00.000Z', 'cm_unit': 'b', 'size': 500, 'filter_charterers': [], 'filter_owners': [], 'filter_products': ['6f11b0724c9a4e85ffa7f1445bc768f054af755a090118dcf99f14745c261653'], 'filter_vessels': [], 'filter_destinations': [], 'filter_origins': ['2d92cc08f22524dba59f6a7e340f132a9da0ce9573cca968eb8e3752ef17a963'], 'filter_storage_locations': [], 'filter_ship_to_ship_locations': [], 'filter_waypoints': [], 'disable_geographic_exclusion_rules': None}\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\r", "Loading from API: 0%| | 0/6948 [00:00\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
cargo_movement_idvessels.0.namevessels.0.imovessels.0.vessel_classvessels.0.dwtvessels.1.namevessels.1.imovessels.1.vessel_classvessels.1.dwtvessels.2.namevessels.2.imovessels.2.vessel_classvessels.2.dwtevents.cargo_port_load_event.0.location.terminal.labelevents.cargo_port_load_event.0.location.port.labelevents.cargo_port_load_event.0.location.country.labelevents.cargo_port_load_event.0.location.trading_region.labelevents.cargo_port_load_event.0.location.shipping_region.labelevents.cargo_port_load_event.1.location.port.labelevents.cargo_port_load_event.2.location.port.labelevents.cargo_port_load_event.0.start_timestampevents.cargo_port_load_event.0.end_timestampevents.cargo_port_load_event.1.end_timestampevents.cargo_port_load_event.2.end_timestampevents.cargo_sts_event.0.start_timestampevents.cargo_sts_event.0.location.sts_zone.labelevents.cargo_sts_event.1.start_timestampevents.cargo_sts_event.1.location.sts_zone.labelevents.cargo_sts_event.2.start_timestampevents.cargo_sts_event.2.location.sts_zone.labelevents.cargo_port_unload_event.0.location.terminal.labelevents.cargo_port_unload_event.0.location.port.labelevents.cargo_port_unload_event.0.location.country.labelevents.cargo_port_unload_event.0.location.region.labelevents.cargo_port_unload_event.0.location.trading_subregion.labelevents.cargo_port_unload_event.0.location.shipping_region.labelevents.cargo_port_unload_event.0.location.trading_region.labelevents.cargo_port_unload_event.0.start_timestampevents.cargo_port_unload_event.0.end_timestampevents.cargo_port_unload_event.1.location.port.labelevents.cargo_port_unload_event.1.end_timestampevents.cargo_port_unload_event.2.location.port.labelevents.cargo_port_unload_event.2.end_timestampvessels.0.corporate_entities.charterer.labelproduct.grade.labelproduct.category.labelproduct.group_product.labelproduct.group.labelquantitystatus
0eef9746529ceeae7acb5568f10ab743cbba3a6400d9a0a29ba66abe2d2e01c36MINERVA ZOE9255684aframax105330NaNNaNNaNNaNNaNNaNNaNNaNPhillips 66 Beaumont TerminalBeaumont, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2018-02-13T17:59:47+00002018-02-18T18:47:19+0000NaNNaNNaNNaNNaNNaNNaNNaNDansk Statoil AsKalundborg [DK]DenmarkEuropeDenmarkNorth SeaNorth West Europe (excl. ARA)2018-03-13T00:11:48+00002018-03-14T06:16:58+0000NaNNaNNaNNaNNaNBakkenLight-SweetCrudeCrude422095unloaded_state
1ef16ba73f8914e697018a9a467c38aeac57cf26f7285f94dded6cc7a2b6334caSONANGOL RANGEL9575541suezmax157755NaNNaNNaNNaNNaNNaNNaNNaNOxy InglesideCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2019-12-09T19:48:16+00002019-12-11T20:44:09+0000NaNNaNNaNNaNNaNNaNNaNNaNCosmo SekiyuSakai [JP]JapanAsiaJapanEast AsiaEast Asia2020-02-18T04:35:00+00002020-02-19T01:23:15+0000NaNNaNNaNNaNCOSMO OILWest Texas Midland (WTM)Light-SweetCrudeCrude138507unloaded_state
2ef1f107c18eb3bdb1d7a16ca9e18e9ca641e68fbb56398a7c1a3dd9dff5f3efeNS CONCEPT9299707aframax109857NaNNaNNaNNaNNaNNaNNaNNaNHouston Enterprise TerminalHouston, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2019-04-12T23:57:05+00002019-04-15T12:30:24+0000NaNNaNNaNNaNNaNNaNNaNNaNOiltanking Puerto BahiaCartagena [CO]ColombiaSouth AmericaNaNSouth AmericaSouth America West Coast2019-04-21T12:23:31+00002019-04-23T22:35:14+0000NaNNaNNaNNaNCHEVRONWest Texas Intermediate (WTI)Light-SweetCrudeCrude567402unloaded_state
3ef2f6505c12546db8624f47a87ab7e9dc280da532905ee8120dc8865c333a126EAGLE TORRANCE9360453aframax107123C. INFINITY9605190.0vlcc_plus313990.0PIPER9282481.0aframax114809.0Nustar North Beach - Corpus ChristiCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2016-09-27T18:00:46+00002016-09-29T16:59:09+0000NaNNaN2016-09-30T12:37:00+0000Southtex Lightering/STS Zone [US]2016-11-18T03:04:13+0000Linggi STS [MY]NaNNaNBP Kwinana RefineryKwinana, WA [AU]AustraliaOceaniaWest Coast AustraliaWest Coast AustraliaOceania2016-12-06T10:20:32+00002016-12-09T10:20:32+0000NaNNaNNaNNaNNaNEagle Ford CondensateNaNCondensatesCrude126217unloaded_state
4ef33bab3fc261e8165bed9796a2b535712322005f26ff85c9f29b176934bce3aMINERVA SOPHIA9382762aframax115873NaNNaNNaNNaNNaNNaNNaNNaNSunoco Logistics Nederland TerminalBeaumont, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2018-07-22T23:08:03+00002018-07-24T11:44:26+0000NaNNaNNaNNaNNaNNaNNaNNaNEsso Fawley Refinery TerminalFawley [GB]United KingdomEuropeEast Coast UKNorth SeaNorth West Europe (excl. ARA)2018-08-10T08:47:45+00002018-08-11T20:31:31+0000NaNNaNNaNNaNNaNWest Texas Midland (WTM)Light-SweetCrudeCrude215312unloaded_state
\n", "" ], "text/plain": [ " cargo_movement_id \\\n", "0 eef9746529ceeae7acb5568f10ab743cbba3a6400d9a0a29ba66abe2d2e01c36 \n", "1 ef16ba73f8914e697018a9a467c38aeac57cf26f7285f94dded6cc7a2b6334ca \n", "2 ef1f107c18eb3bdb1d7a16ca9e18e9ca641e68fbb56398a7c1a3dd9dff5f3efe \n", "3 ef2f6505c12546db8624f47a87ab7e9dc280da532905ee8120dc8865c333a126 \n", "4 ef33bab3fc261e8165bed9796a2b535712322005f26ff85c9f29b176934bce3a \n", "\n", " vessels.0.name vessels.0.imo vessels.0.vessel_class vessels.0.dwt \\\n", "0 MINERVA ZOE 9255684 aframax 105330 \n", "1 SONANGOL RANGEL 9575541 suezmax 157755 \n", "2 NS CONCEPT 9299707 aframax 109857 \n", "3 EAGLE TORRANCE 9360453 aframax 107123 \n", "4 MINERVA SOPHIA 9382762 aframax 115873 \n", "\n", " vessels.1.name vessels.1.imo vessels.1.vessel_class vessels.1.dwt \\\n", "0 NaN NaN NaN NaN \n", "1 NaN NaN NaN NaN \n", "2 NaN NaN NaN NaN \n", "3 C. INFINITY 9605190.0 vlcc_plus 313990.0 \n", "4 NaN NaN NaN NaN \n", "\n", " vessels.2.name vessels.2.imo vessels.2.vessel_class vessels.2.dwt \\\n", "0 NaN NaN NaN NaN \n", "1 NaN NaN NaN NaN \n", "2 NaN NaN NaN NaN \n", "3 PIPER 9282481.0 aframax 114809.0 \n", "4 NaN NaN NaN NaN \n", "\n", " events.cargo_port_load_event.0.location.terminal.label \\\n", "0 Phillips 66 Beaumont Terminal \n", "1 Oxy Ingleside \n", "2 Houston Enterprise Terminal \n", "3 Nustar North Beach - Corpus Christi \n", "4 Sunoco Logistics Nederland Terminal \n", "\n", " events.cargo_port_load_event.0.location.port.label \\\n", "0 Beaumont, TX [US] \n", "1 Corpus Christi, TX [US] \n", "2 Houston, TX [US] \n", "3 Corpus Christi, TX [US] \n", "4 Beaumont, TX [US] \n", "\n", " events.cargo_port_load_event.0.location.country.label \\\n", "0 United States \n", "1 United States \n", "2 United States \n", "3 United States \n", "4 United States \n", "\n", " events.cargo_port_load_event.0.location.trading_region.label \\\n", "0 PADD 3 (US Gulf Coast) \n", "1 PADD 3 (US Gulf Coast) \n", "2 PADD 3 (US Gulf Coast) \n", "3 PADD 3 (US Gulf Coast) \n", "4 PADD 3 (US Gulf Coast) \n", "\n", " events.cargo_port_load_event.0.location.shipping_region.label \\\n", "0 USAC/USGC \n", "1 USAC/USGC \n", "2 USAC/USGC \n", "3 USAC/USGC \n", "4 USAC/USGC \n", "\n", " events.cargo_port_load_event.1.location.port.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_load_event.2.location.port.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_load_event.0.start_timestamp \\\n", "0 2018-02-13T17:59:47+0000 \n", "1 2019-12-09T19:48:16+0000 \n", "2 2019-04-12T23:57:05+0000 \n", "3 2016-09-27T18:00:46+0000 \n", "4 2018-07-22T23:08:03+0000 \n", "\n", " events.cargo_port_load_event.0.end_timestamp \\\n", "0 2018-02-18T18:47:19+0000 \n", "1 2019-12-11T20:44:09+0000 \n", "2 2019-04-15T12:30:24+0000 \n", "3 2016-09-29T16:59:09+0000 \n", "4 2018-07-24T11:44:26+0000 \n", "\n", " events.cargo_port_load_event.1.end_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_load_event.2.end_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_sts_event.0.start_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 2016-09-30T12:37:00+0000 \n", "4 NaN \n", "\n", " events.cargo_sts_event.0.location.sts_zone.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 Southtex Lightering/STS Zone [US] \n", "4 NaN \n", "\n", " events.cargo_sts_event.1.start_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 2016-11-18T03:04:13+0000 \n", "4 NaN \n", "\n", " events.cargo_sts_event.1.location.sts_zone.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 Linggi STS [MY] \n", "4 NaN \n", "\n", " events.cargo_sts_event.2.start_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_sts_event.2.location.sts_zone.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_unload_event.0.location.terminal.label \\\n", "0 Dansk Statoil As \n", "1 Cosmo Sekiyu \n", "2 Oiltanking Puerto Bahia \n", "3 BP Kwinana Refinery \n", "4 Esso Fawley Refinery Terminal \n", "\n", " events.cargo_port_unload_event.0.location.port.label \\\n", "0 Kalundborg [DK] \n", "1 Sakai [JP] \n", "2 Cartagena [CO] \n", "3 Kwinana, WA [AU] \n", "4 Fawley [GB] \n", "\n", " events.cargo_port_unload_event.0.location.country.label \\\n", "0 Denmark \n", "1 Japan \n", "2 Colombia \n", "3 Australia \n", "4 United Kingdom \n", "\n", " events.cargo_port_unload_event.0.location.region.label \\\n", "0 Europe \n", "1 Asia \n", "2 South America \n", "3 Oceania \n", "4 Europe \n", "\n", " events.cargo_port_unload_event.0.location.trading_subregion.label \\\n", "0 Denmark \n", "1 Japan \n", "2 NaN \n", "3 West Coast Australia \n", "4 East Coast UK \n", "\n", " events.cargo_port_unload_event.0.location.shipping_region.label \\\n", "0 North Sea \n", "1 East Asia \n", "2 South America \n", "3 West Coast Australia \n", "4 North Sea \n", "\n", " events.cargo_port_unload_event.0.location.trading_region.label \\\n", "0 North West Europe (excl. ARA) \n", "1 East Asia \n", "2 South America West Coast \n", "3 Oceania \n", "4 North West Europe (excl. ARA) \n", "\n", " events.cargo_port_unload_event.0.start_timestamp \\\n", "0 2018-03-13T00:11:48+0000 \n", "1 2020-02-18T04:35:00+0000 \n", "2 2019-04-21T12:23:31+0000 \n", "3 2016-12-06T10:20:32+0000 \n", "4 2018-08-10T08:47:45+0000 \n", "\n", " events.cargo_port_unload_event.0.end_timestamp \\\n", "0 2018-03-14T06:16:58+0000 \n", "1 2020-02-19T01:23:15+0000 \n", "2 2019-04-23T22:35:14+0000 \n", "3 2016-12-09T10:20:32+0000 \n", "4 2018-08-11T20:31:31+0000 \n", "\n", " events.cargo_port_unload_event.1.location.port.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_unload_event.1.end_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_unload_event.2.location.port.label \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " events.cargo_port_unload_event.2.end_timestamp \\\n", "0 NaN \n", "1 NaN \n", "2 NaN \n", "3 NaN \n", "4 NaN \n", "\n", " vessels.0.corporate_entities.charterer.label product.grade.label \\\n", "0 NaN Bakken \n", "1 COSMO OIL West Texas Midland (WTM) \n", "2 CHEVRON West Texas Intermediate (WTI) \n", "3 NaN Eagle Ford Condensate \n", "4 NaN West Texas Midland (WTM) \n", "\n", " product.category.label product.group_product.label product.group.label \\\n", "0 Light-Sweet Crude Crude \n", "1 Light-Sweet Crude Crude \n", "2 Light-Sweet Crude Crude \n", "3 NaN Condensates Crude \n", "4 Light-Sweet Crude Crude \n", "\n", " quantity status \n", "0 422095 unloaded_state \n", "1 138507 unloaded_state \n", "2 567402 unloaded_state \n", "3 126217 unloaded_state \n", "4 215312 unloaded_state " ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Query the SDK and convert result to DataFrame\n", "cms = CargoMovements().search(\n", " filter_activity = 'loading_end',\n", " filter_origins = us,\n", " filter_products = crude,\n", " filter_time_min = START_DATE,\n", " filter_time_max = END_DATE,\n", " cm_unit = 'b'\n", " ).to_df(columns = required_columns)\n", "print('Fetched {} crude cargo movements from United States'.format(cms.shape[0]))\n", "cms.head()" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "4,618 crude movements remaining after removing intra-movements\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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
cargo_movement_idvessels.0.namevessels.0.imovessels.0.vessel_classvessels.0.dwtvessels.1.namevessels.1.imovessels.1.vessel_classvessels.1.dwtvessels.2.namevessels.2.imovessels.2.vessel_classvessels.2.dwtevents.cargo_port_load_event.0.location.terminal.labelloading_portloading_countryloading_trading_regionevents.cargo_port_load_event.0.location.shipping_region.labelevents.cargo_port_load_event.1.location.port.labelevents.cargo_port_load_event.2.location.port.labelloading_timestampstart_timestampevents.cargo_port_load_event.1.end_timestampevents.cargo_port_load_event.2.end_timestampevents.cargo_sts_event.0.start_timestampevents.cargo_sts_event.0.location.sts_zone.labelevents.cargo_sts_event.1.start_timestampevents.cargo_sts_event.1.location.sts_zone.labelevents.cargo_sts_event.2.start_timestampevents.cargo_sts_event.2.location.sts_zone.labelevents.cargo_port_unload_event.0.location.terminal.labelevents.cargo_port_unload_event.0.location.port.labelunloading_countryevents.cargo_port_unload_event.0.location.region.labelevents.cargo_port_unload_event.0.location.trading_subregion.labelevents.cargo_port_unload_event.0.location.shipping_region.labelevents.cargo_port_unload_event.0.location.trading_region.labelunloading_timestampend_timestampevents.cargo_port_unload_event.1.location.port.labelevents.cargo_port_unload_event.1.end_timestampevents.cargo_port_unload_event.2.location.port.labelevents.cargo_port_unload_event.2.end_timestampvessels.0.corporate_entities.charterer.labelproduct.grade.labelproduct_categoryproduct.group_product.labelproduct.group.labelquantitystatusloading_weekloading_month
60285fb183365a5f497db067b7f90c3974a372960380abb40a49da0a4769dd86a1e2GARIBALDI SPIRIT9422835aframax109039DRENEC9723100.0vlcc_plus299999.0NaNNaNNaNNaNBuckeye TerminalCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2017-09-21 15:50:342017-09-24 12:35:17NaTNaT2017-09-25 12:00:31Southtex Lightering/STS Zone [US]NaTNaNNaTNaNVadinar IOC SBMVadinar [IN]IndiaAsiaWest Coast IndiaWest Coast IndiaIndian Sub-Continent2017-11-10 09:22:512017-11-12 15:21:46NaNNaTNaNNaTNaNEagle Ford crudeLight-SweetCrudeCrude150464unloaded_state2017-09-182017-09-01
2759e7d3a0a3f4d59de7a7765a65c13e166171564e61823dadf315a67cf873e09571EUROVISION9567697suezmax157803NaNNaNNaNNaNNaNNaNNaNNaNNustar North Beach - Corpus ChristiCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2019-09-21 17:33:542019-09-25 19:50:44NaTNaTNaTNaNNaTNaNNaTNaNOilhub Korea Yeosu (OKYC)Yeosu (Yosu), Gwangyang [KR]South KoreaAsiaSouth KoreaEast AsiaEast Asia2019-11-19 13:14:042019-11-19 13:35:23NaNNaTNaNNaTSHELLEagle Ford crudeLight-SweetCrudeCrude311684unloaded_state2019-09-232019-09-01
126f39303edd4e4a80c0e79eaa57f19677bbc38dfb607a7ded67bfe16f7665cc5b0NIKOLAY ZUYEV9610781suezmax122039SAHBA9388273.0vlcc_plus317563.0NaNNaNNaNNaNBuckeye TerminalCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2018-06-17 21:58:422018-06-20 15:01:53NaTNaT2018-07-06 09:39:00Southtex Lightering/STS Zone [US]NaTNaNNaTNaNVadinar IOC SBMVadinar [IN]IndiaAsiaWest Coast IndiaWest Coast IndiaIndian Sub-Continent2018-08-20 03:05:182018-08-22 04:14:36NaNNaTNaNNaTNaNLight Louisiana Sweet (LLS)Light-SweetCrudeCrude877175unloaded_state2018-06-182018-06-01
194f6393679a7a7af212332505c15df207595aae9d0bd75de54116f108b429142afPARAMOUNT HELSINKI9453963aframax114165DORRA9386964.0vlcc_plus317458.0NaNNaNNaNNaNOxy InglesideCorpus Christi, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2018-08-05 18:48:332018-08-07 12:35:13NaTNaT2018-08-08 12:25:09Southtex Lightering/STS Zone [US]NaTNaNNaTNaNCpc - ShalungShalung [TW]TaiwanAsiaTaiwanEast AsiaEast Asia2018-10-09 00:43:132018-10-09 21:44:13NaNNaTNaNNaTNaNWest Texas Intermediate (WTI)Light-SweetCrudeCrude270492unloaded_state2018-08-062018-08-01
687147fdf37ae0f374c535be9c9f72d48ef2a313e7a79c2312d5ae5818cbd62dd171RUNNER9749518suezmax158594NaNNaNNaNNaNNaNNaNNaNNaNSunoco Logistics Nederland TerminalBeaumont, TX [US]United StatesPADD 3 (US Gulf Coast)USAC/USGCNaNNaN2020-01-19 21:28:342020-01-20 20:16:25NaTNaTNaTNaNNaTNaNNaTNaNNaNCilacap, Java [ID]IndonesiaAsiaIndonesiaFar EastSoutheast Asia2020-03-09 05:18:01NaTNaNNaTNaNNaTTOTALWest Texas Intermediate (WTI)Light-SweetCrudeCrude321761transiting_state2020-01-202020-01-01
\n", "
" ], "text/plain": [ " cargo_movement_id \\\n", "6028 5fb183365a5f497db067b7f90c3974a372960380abb40a49da0a4769dd86a1e2 \n", "2759 e7d3a0a3f4d59de7a7765a65c13e166171564e61823dadf315a67cf873e09571 \n", "126 f39303edd4e4a80c0e79eaa57f19677bbc38dfb607a7ded67bfe16f7665cc5b0 \n", "194 f6393679a7a7af212332505c15df207595aae9d0bd75de54116f108b429142af \n", "6871 47fdf37ae0f374c535be9c9f72d48ef2a313e7a79c2312d5ae5818cbd62dd171 \n", "\n", " vessels.0.name vessels.0.imo vessels.0.vessel_class vessels.0.dwt \\\n", "6028 GARIBALDI SPIRIT 9422835 aframax 109039 \n", "2759 EUROVISION 9567697 suezmax 157803 \n", "126 NIKOLAY ZUYEV 9610781 suezmax 122039 \n", "194 PARAMOUNT HELSINKI 9453963 aframax 114165 \n", "6871 RUNNER 9749518 suezmax 158594 \n", "\n", " vessels.1.name vessels.1.imo vessels.1.vessel_class vessels.1.dwt \\\n", "6028 DRENEC 9723100.0 vlcc_plus 299999.0 \n", "2759 NaN NaN NaN NaN \n", "126 SAHBA 9388273.0 vlcc_plus 317563.0 \n", "194 DORRA 9386964.0 vlcc_plus 317458.0 \n", "6871 NaN NaN NaN NaN \n", "\n", " vessels.2.name vessels.2.imo vessels.2.vessel_class vessels.2.dwt \\\n", "6028 NaN NaN NaN NaN \n", "2759 NaN NaN NaN NaN \n", "126 NaN NaN NaN NaN \n", "194 NaN NaN NaN NaN \n", "6871 NaN NaN NaN NaN \n", "\n", " events.cargo_port_load_event.0.location.terminal.label \\\n", "6028 Buckeye Terminal \n", "2759 Nustar North Beach - Corpus Christi \n", "126 Buckeye Terminal \n", "194 Oxy Ingleside \n", "6871 Sunoco Logistics Nederland Terminal \n", "\n", " loading_port loading_country loading_trading_region \\\n", "6028 Corpus Christi, TX [US] United States PADD 3 (US Gulf Coast) \n", "2759 Corpus Christi, TX [US] United States PADD 3 (US Gulf Coast) \n", "126 Corpus Christi, TX [US] United States PADD 3 (US Gulf Coast) \n", "194 Corpus Christi, TX [US] United States PADD 3 (US Gulf Coast) \n", "6871 Beaumont, TX [US] United States PADD 3 (US Gulf Coast) \n", "\n", " events.cargo_port_load_event.0.location.shipping_region.label \\\n", "6028 USAC/USGC \n", "2759 USAC/USGC \n", "126 USAC/USGC \n", "194 USAC/USGC \n", "6871 USAC/USGC \n", "\n", " events.cargo_port_load_event.1.location.port.label \\\n", "6028 NaN \n", "2759 NaN \n", "126 NaN \n", "194 NaN \n", "6871 NaN \n", "\n", " events.cargo_port_load_event.2.location.port.label loading_timestamp \\\n", "6028 NaN 2017-09-21 15:50:34 \n", "2759 NaN 2019-09-21 17:33:54 \n", "126 NaN 2018-06-17 21:58:42 \n", "194 NaN 2018-08-05 18:48:33 \n", "6871 NaN 2020-01-19 21:28:34 \n", "\n", " start_timestamp events.cargo_port_load_event.1.end_timestamp \\\n", "6028 2017-09-24 12:35:17 NaT \n", "2759 2019-09-25 19:50:44 NaT \n", "126 2018-06-20 15:01:53 NaT \n", "194 2018-08-07 12:35:13 NaT \n", "6871 2020-01-20 20:16:25 NaT \n", "\n", " events.cargo_port_load_event.2.end_timestamp \\\n", "6028 NaT \n", "2759 NaT \n", "126 NaT \n", "194 NaT \n", "6871 NaT \n", "\n", " events.cargo_sts_event.0.start_timestamp \\\n", "6028 2017-09-25 12:00:31 \n", "2759 NaT \n", "126 2018-07-06 09:39:00 \n", "194 2018-08-08 12:25:09 \n", "6871 NaT \n", "\n", " events.cargo_sts_event.0.location.sts_zone.label \\\n", "6028 Southtex Lightering/STS Zone [US] \n", "2759 NaN \n", "126 Southtex Lightering/STS Zone [US] \n", "194 Southtex Lightering/STS Zone [US] \n", "6871 NaN \n", "\n", " events.cargo_sts_event.1.start_timestamp \\\n", "6028 NaT \n", "2759 NaT \n", "126 NaT \n", "194 NaT \n", "6871 NaT \n", "\n", " events.cargo_sts_event.1.location.sts_zone.label \\\n", "6028 NaN \n", "2759 NaN \n", "126 NaN \n", "194 NaN \n", "6871 NaN \n", "\n", " events.cargo_sts_event.2.start_timestamp \\\n", "6028 NaT \n", "2759 NaT \n", "126 NaT \n", "194 NaT \n", "6871 NaT \n", "\n", " events.cargo_sts_event.2.location.sts_zone.label \\\n", "6028 NaN \n", "2759 NaN \n", "126 NaN \n", "194 NaN \n", "6871 NaN \n", "\n", " events.cargo_port_unload_event.0.location.terminal.label \\\n", "6028 Vadinar IOC SBM \n", "2759 Oilhub Korea Yeosu (OKYC) \n", "126 Vadinar IOC SBM \n", "194 Cpc - Shalung \n", "6871 NaN \n", "\n", " events.cargo_port_unload_event.0.location.port.label unloading_country \\\n", "6028 Vadinar [IN] India \n", "2759 Yeosu (Yosu), Gwangyang [KR] South Korea \n", "126 Vadinar [IN] India \n", "194 Shalung [TW] Taiwan \n", "6871 Cilacap, Java [ID] Indonesia \n", "\n", " events.cargo_port_unload_event.0.location.region.label \\\n", "6028 Asia \n", "2759 Asia \n", "126 Asia \n", "194 Asia \n", "6871 Asia \n", "\n", " events.cargo_port_unload_event.0.location.trading_subregion.label \\\n", "6028 West Coast India \n", "2759 South Korea \n", "126 West Coast India \n", "194 Taiwan \n", "6871 Indonesia \n", "\n", " events.cargo_port_unload_event.0.location.shipping_region.label \\\n", "6028 West Coast India \n", "2759 East Asia \n", "126 West Coast India \n", "194 East Asia \n", "6871 Far East \n", "\n", " events.cargo_port_unload_event.0.location.trading_region.label \\\n", "6028 Indian Sub-Continent \n", "2759 East Asia \n", "126 Indian Sub-Continent \n", "194 East Asia \n", "6871 Southeast Asia \n", "\n", " unloading_timestamp end_timestamp \\\n", "6028 2017-11-10 09:22:51 2017-11-12 15:21:46 \n", "2759 2019-11-19 13:14:04 2019-11-19 13:35:23 \n", "126 2018-08-20 03:05:18 2018-08-22 04:14:36 \n", "194 2018-10-09 00:43:13 2018-10-09 21:44:13 \n", "6871 2020-03-09 05:18:01 NaT \n", "\n", " events.cargo_port_unload_event.1.location.port.label \\\n", "6028 NaN \n", "2759 NaN \n", "126 NaN \n", "194 NaN \n", "6871 NaN \n", "\n", " events.cargo_port_unload_event.1.end_timestamp \\\n", "6028 NaT \n", "2759 NaT \n", "126 NaT \n", "194 NaT \n", "6871 NaT \n", "\n", " events.cargo_port_unload_event.2.location.port.label \\\n", "6028 NaN \n", "2759 NaN \n", "126 NaN \n", "194 NaN \n", "6871 NaN \n", "\n", " events.cargo_port_unload_event.2.end_timestamp \\\n", "6028 NaT \n", "2759 NaT \n", "126 NaT \n", "194 NaT \n", "6871 NaT \n", "\n", " vessels.0.corporate_entities.charterer.label \\\n", "6028 NaN \n", "2759 SHELL \n", "126 NaN \n", "194 NaN \n", "6871 TOTAL \n", "\n", " product.grade.label product_category \\\n", "6028 Eagle Ford crude Light-Sweet \n", "2759 Eagle Ford crude Light-Sweet \n", "126 Light Louisiana Sweet (LLS) Light-Sweet \n", "194 West Texas Intermediate (WTI) Light-Sweet \n", "6871 West Texas Intermediate (WTI) Light-Sweet \n", "\n", " product.group_product.label product.group.label quantity \\\n", "6028 Crude Crude 150464 \n", "2759 Crude Crude 311684 \n", "126 Crude Crude 877175 \n", "194 Crude Crude 270492 \n", "6871 Crude Crude 321761 \n", "\n", " status loading_week loading_month \n", "6028 unloaded_state 2017-09-18 2017-09-01 \n", "2759 unloaded_state 2019-09-23 2019-09-01 \n", "126 unloaded_state 2018-06-18 2018-06-01 \n", "194 unloaded_state 2018-08-06 2018-08-01 \n", "6871 transiting_state 2020-01-20 2020-01-01 " ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Perform some basic operations in the DataFrame\n", "cms = prepare_cms(cms, ignore_intra = True)\n", "print('{:,} crude movements remaining after removing intra-movements'.format(cms.shape[0]))\n", "cms.sample(5)" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Minimum start timestamp: 2016-02-02 13:04:45 \n", "Maximum start timestamp: 2020-01-31 21:14:48\n" ] } ], "source": [ "# Check minimum and maximum start_timestamps\n", "print('Minimum start timestamp: {} \\nMaximum start timestamp: {}'.\\\n", " format(cms['start_timestamp'].min(), cms['start_timestamp'].max()))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Having fetched our data, we are now ready to start exploring them and see what kind of questions we can answer." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Monthly US Exports" ] }, { "cell_type": "code", "execution_count": 11, "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", "
barrels
loading_month
2019-04-0177648792
2019-05-0180818763
2019-06-0192063064
2019-07-0176972694
2019-08-0179035290
2019-09-0191128679
2019-10-01100686490
2019-11-0179899088
2019-12-01103536412
2020-01-0191491919
\n", "
" ], "text/plain": [ " barrels\n", "loading_month \n", "2019-04-01 77648792\n", "2019-05-01 80818763\n", "2019-06-01 92063064\n", "2019-07-01 76972694\n", "2019-08-01 79035290\n", "2019-09-01 91128679\n", "2019-10-01 100686490\n", "2019-11-01 79899088\n", "2019-12-01 103536412\n", "2020-01-01 91491919" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "monthly_quantity = cms.groupby('loading_month').\\\n", " agg({'quantity': 'sum'}).\\\n", " rename(columns = {'quantity': 'barrels'})\n", "\n", "monthly_quantity.tail(10)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 0, 'Month')" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "ax = monthly_quantity.plot(kind='bar', \n", " figsize = (15,5), \n", " rot=60,\n", " title = 'US Monthly Crude Exports')\n", "ax.set_ylabel('Quantity (barrels)')\n", "ax.set_xlabel('Month')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Grades Breakdown" ] }, { "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", "
product_categoryHeavy-SourLight-SourLight-SweetMedium-SourMedium-Sweet
loading_month
2016-02-010.00.06365508.00.00.0
2016-03-010.00.09770342.00.0550011.0
2016-04-010.00.08568627.00.0397395.0
2016-05-010.00.010014700.0986121.00.0
2016-06-010.00.04516795.0435939.0142074.0
\n", "
" ], "text/plain": [ "product_category Heavy-Sour Light-Sour Light-Sweet Medium-Sour \\\n", "loading_month \n", "2016-02-01 0.0 0.0 6365508.0 0.0 \n", "2016-03-01 0.0 0.0 9770342.0 0.0 \n", "2016-04-01 0.0 0.0 8568627.0 0.0 \n", "2016-05-01 0.0 0.0 10014700.0 986121.0 \n", "2016-06-01 0.0 0.0 4516795.0 435939.0 \n", "\n", "product_category Medium-Sweet \n", "loading_month \n", "2016-02-01 0.0 \n", "2016-03-01 550011.0 \n", "2016-04-01 397395.0 \n", "2016-05-01 0.0 \n", "2016-06-01 142074.0 " ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "quantity_by_category = cms.groupby(by = ['loading_month','product_category']).\\\n", " agg({'quantity': 'sum'}).\\\n", " rename(columns = {'quantity': 'barrels'}).\\\n", " reset_index()\n", "quantity_by_category = quantity_by_category.pivot(index = 'loading_month', \n", " columns = 'product_category',\n", " values = 'barrels')\n", "quantity_by_category = quantity_by_category.fillna(0)\n", "quantity_by_category.head()" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 0, 'Month')" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "ax2 = quantity_by_category.plot.bar(stacked=True, \n", " figsize = (12,5), \n", " rot = 60,\n", " title = 'US Crude Exports \\n Breakdown by Sulphur Content & API')\n", "ax2.set_ylabel('Quantity (barrels)')\n", "ax2.set_xlabel('Month')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Top Destinations" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Who are the main **receivers** of US crude starting from 2019 onwards?" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "scrolled": true }, "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", "
quantityperc
South Korea15390598014.28
Canada12767469111.85
United Kingdom1047326639.72
Netherlands1016359419.43
India889446468.26
Taiwan525284744.88
Italy492387414.57
China417090183.87
France410326953.81
Thailand300404712.79
Other28598883026.54
\n", "
" ], "text/plain": [ " quantity perc\n", "South Korea 153905980 14.28\n", "Canada 127674691 11.85\n", "United Kingdom 104732663 9.72\n", "Netherlands 101635941 9.43\n", "India 88944646 8.26\n", "Taiwan 52528474 4.88\n", "Italy 49238741 4.57\n", "China 41709018 3.87\n", "France 41032695 3.81\n", "Thailand 30040471 2.79\n", "Other 285988830 26.54" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "NUM_TOP_DESTINATIONS = 10\n", "\n", "# Filter all movements taht started from 2019 onwards\n", "cms_2019 = cms.loc[cms['start_timestamp'] >= '2019-01-01']\n", "\n", "# Calculate quantity per destination\n", "quantity_per_destination = cms_2019.groupby('unloading_country').\\\n", " agg({'quantity': 'sum'}).\\\n", " sort_values(by = 'quantity', ascending = False)\n", "\n", "# Select top destinations and group together the remaining ones\n", "top_destination_countries = quantity_per_destination.head(NUM_TOP_DESTINATIONS)\n", "rest = pd.DataFrame(index = ['Other'], columns = ['quantity'])\n", "rest.loc['Other'] = quantity_per_destination[NUM_TOP_DESTINATIONS:].sum().values\n", "top_destination_countries = pd.concat([top_destination_countries, rest])\n", "top_destination_countries['perc'] = round(top_destination_countries['quantity']*100 / top_destination_countries['quantity'].sum(),2)\n", "\n", "display(top_destination_countries)" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Pick a colormap pallette and generate random colors\n", "vals = np.linspace(0,1,top_destination_countries.shape[0])\n", "np.random.shuffle(vals)\n", "cmap = plt.cm.colors.ListedColormap(plt.cm.Greens(vals))\n", "\n", "plt.figure(figsize = (10,5))\n", "plt.pie(top_destination_countries['quantity'],\n", " labels = top_destination_countries.index.values,\n", " shadow = False,\n", " colors = cmap.colors,\n", " startangle=90,\n", " autopct = '%1.2f%%')\n", "plt.title('Top US Crude Destinations', fontsize=13)\n", "plt.axis('equal')\n", "plt.tight_layout()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Simple Forecasting Model" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Having done an initial exploration of the US crude exports, now is time to make it even more interesing and see how an analyst or data scientist could use these data to build a predictive model. For this purpose, we will try to predict **future** flows using an open source library called **Prophet** that was developed by Facebook. Although there is no free-lunch, Prophet is a very intuitive library, offers quick and handy results and is ideal for an initial iteration / baseline models on a wide variety of forecasting problems.\n", "\n", "It is based on a `Generalized Additive Model (GAM)` with three main components: `trend`, `seasonality` and `holidays`:\n", "$$ y(t) = g(t) + s(t) + h(t) + \\epsilon_{t}$$\n", "\n", "For the purposes of this notebook, we will use the default parameters to fit our model. In order to install Prophet, you can simply run `pip install fbprophet` or `conda install -c conda-forge fbprophet` for conda users." ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "ERROR:fbprophet:Importing plotly failed. Interactive plots will not work.\n" ] } ], "source": [ "from fbprophet import Prophet\n", "from fbprophet.plot import add_changepoints_to_plot\n", "from fbprophet.diagnostics import cross_validation\n", "from fbprophet.diagnostics import performance_metrics\n", "from fbprophet.plot import plot_cross_validation_metric" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Firstly, we need to prepare our DataFrame. The input to Prophet must always be a DataFrame with two columns: `ds` (datestamp in Pandas expected format) and `y` (numerical value to be predicted). " ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "scrolled": true }, "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", "
dsy
432019-09-0191128679
442019-10-01100686490
452019-11-0179899088
462019-12-01103536412
472020-01-0191491919
\n", "
" ], "text/plain": [ " ds y\n", "43 2019-09-01 91128679\n", "44 2019-10-01 100686490\n", "45 2019-11-01 79899088\n", "46 2019-12-01 103536412\n", "47 2020-01-01 91491919" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = monthly_quantity.copy().reset_index()\n", "df = df.rename(columns = {'loading_month': 'ds', 'barrels': 'y'})\n", "df.tail()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Model Fitting and Cross-Validation" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can now initialize and fit our model. Since we will deal with monthly data we explicitly ignore weekly seasonality." ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "INFO:fbprophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.\n" ] }, { "data": { "text/plain": [ "" ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "m = Prophet(weekly_seasonality=False)\n", "m.fit(df)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Prophet offers a built-in method for cross-validation using *simulated historical forecasts (SHFs)*. What this essentially does is to select some *cutoff* points in the history, for each of them fit a model using data only up to the cutoff point and then make predictions for a given forecast horizon. The user needs to specify the `horizon (H)` and optionally the size of the `initial` training period and the `period` (i.e. spacing) between the cutoff dates. As a heuristic, for a given horizon `H` it is suggested that a simulated forecast runs every `H/2` periods. Since we are dealing with monthly data we will set our initial training size to 2 years (24 months) to be able to account for yearly seasonality, our forecasted horizon to 6 months and the period to 3 months.\n", "\n", "One important thing to notice is the following: Prophet's build-in `cross_validation` method accepts period arguments that are **timedelta** objects. In this notebook, we are dealing with monthly data, however, pandas latest version (i.e. **1.0.1**) doesn't seem to work with a month unit in the `pd.to_timedelta` method (although pandas docs mention that an `M` option is supported as a unit, the following error pops-up when trying to run: *ValueError: Units 'M' and 'Y' are no longer supported, as they do not represent unambiguous timedelta values durations*). A work-around through this is to define periods in days and multiply by the **average number of days per month**, as defined in the Gregorian calendar (i.e. 30.436875). With this conversion the results will be identical with what `pd.to_timedelta()` with `unit=''M` would produce in older pandas versions." ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "scrolled": true }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "INFO:fbprophet:Making 6 forecasts with cutoffs between 2018-04-01 19:48:54 and 2019-07-02 09:05:24\n", "INFO:fbprophet:n_changepoints greater than number of observations.Using 20.\n", "INFO:fbprophet:n_changepoints greater than number of observations.Using 23.\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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
dsyhatyhat_loweryhat_upperycutoff
02018-05-015.040042e+074.967177e+075.112307e+07513967202018-04-01
12018-06-013.664930e+073.587002e+073.751328e+07624846122018-04-01
22018-07-014.853681e+074.763286e+074.957673e+07562894652018-04-01
32018-08-013.523361e+073.413859e+073.646861e+07543228802018-04-01
42018-09-015.316111e+075.186468e+075.469529e+07581027682018-04-01
52018-10-018.382423e+078.210670e+078.572572e+07662069192018-04-01
62018-08-013.750728e+073.495334e+074.019379e+07543228802018-07-02
72018-09-015.491614e+075.214068e+075.755979e+07581027682018-07-02
82018-10-018.613946e+078.326978e+078.882009e+07662069192018-07-02
92018-11-015.251827e+074.998365e+075.516644e+07642471002018-07-02
102018-12-017.083287e+076.788084e+077.372896e+07651150132018-07-02
112018-11-015.185985e+074.794188e+075.596643e+07642471002018-10-01
122018-12-017.152869e+076.741776e+077.568692e+07651150132018-10-01
132019-01-016.278689e+075.860127e+076.698408e+07777215492018-10-01
142019-02-017.387505e+076.976284e+077.795319e+07693611142018-10-01
152019-03-017.378256e+076.939656e+077.802458e+07816229382018-10-01
162019-04-017.982540e+077.595450e+078.389104e+07776487922018-10-01
172019-01-016.441183e+076.039529e+076.833383e+07777215492018-12-31
182019-02-017.589843e+077.205305e+077.973862e+07693611142018-12-31
192019-03-017.598118e+077.211916e+078.001506e+07816229382018-12-31
202019-04-018.226506e+077.809231e+078.618986e+07776487922018-12-31
212019-05-018.044857e+077.628468e+078.434376e+07808187632018-12-31
222019-06-019.425433e+079.040677e+079.820845e+07920630642018-12-31
232019-07-018.644319e+078.209196e+079.028962e+07769726942018-12-31
242019-05-018.133520e+077.723318e+078.567094e+07808187632019-04-02
252019-06-019.609183e+079.203950e+071.001646e+08920630642019-04-02
262019-07-018.867596e+078.443326e+079.269204e+07769726942019-04-02
272019-08-018.127349e+077.731661e+078.532102e+07790352902019-04-02
282019-09-018.582029e+078.163789e+078.968897e+07911286792019-04-02
292019-10-011.034187e+089.929004e+071.076682e+081006864902019-04-02
302019-08-018.131430e+077.717230e+078.549134e+07790352902019-07-02
312019-09-018.525117e+078.106940e+078.974195e+07911286792019-07-02
322019-10-011.020143e+089.796074e+071.062698e+081006864902019-07-02
332019-11-019.165255e+078.748693e+079.568568e+07798990882019-07-02
342019-12-019.346174e+078.930227e+079.781523e+071035364122019-07-02
352020-01-011.055244e+081.011898e+081.097860e+08914919192019-07-02
\n", "
" ], "text/plain": [ " ds yhat yhat_lower yhat_upper y cutoff\n", "0 2018-05-01 5.040042e+07 4.967177e+07 5.112307e+07 51396720 2018-04-01\n", "1 2018-06-01 3.664930e+07 3.587002e+07 3.751328e+07 62484612 2018-04-01\n", "2 2018-07-01 4.853681e+07 4.763286e+07 4.957673e+07 56289465 2018-04-01\n", "3 2018-08-01 3.523361e+07 3.413859e+07 3.646861e+07 54322880 2018-04-01\n", "4 2018-09-01 5.316111e+07 5.186468e+07 5.469529e+07 58102768 2018-04-01\n", "5 2018-10-01 8.382423e+07 8.210670e+07 8.572572e+07 66206919 2018-04-01\n", "6 2018-08-01 3.750728e+07 3.495334e+07 4.019379e+07 54322880 2018-07-02\n", "7 2018-09-01 5.491614e+07 5.214068e+07 5.755979e+07 58102768 2018-07-02\n", "8 2018-10-01 8.613946e+07 8.326978e+07 8.882009e+07 66206919 2018-07-02\n", "9 2018-11-01 5.251827e+07 4.998365e+07 5.516644e+07 64247100 2018-07-02\n", "10 2018-12-01 7.083287e+07 6.788084e+07 7.372896e+07 65115013 2018-07-02\n", "11 2018-11-01 5.185985e+07 4.794188e+07 5.596643e+07 64247100 2018-10-01\n", "12 2018-12-01 7.152869e+07 6.741776e+07 7.568692e+07 65115013 2018-10-01\n", "13 2019-01-01 6.278689e+07 5.860127e+07 6.698408e+07 77721549 2018-10-01\n", "14 2019-02-01 7.387505e+07 6.976284e+07 7.795319e+07 69361114 2018-10-01\n", "15 2019-03-01 7.378256e+07 6.939656e+07 7.802458e+07 81622938 2018-10-01\n", "16 2019-04-01 7.982540e+07 7.595450e+07 8.389104e+07 77648792 2018-10-01\n", "17 2019-01-01 6.441183e+07 6.039529e+07 6.833383e+07 77721549 2018-12-31\n", "18 2019-02-01 7.589843e+07 7.205305e+07 7.973862e+07 69361114 2018-12-31\n", "19 2019-03-01 7.598118e+07 7.211916e+07 8.001506e+07 81622938 2018-12-31\n", "20 2019-04-01 8.226506e+07 7.809231e+07 8.618986e+07 77648792 2018-12-31\n", "21 2019-05-01 8.044857e+07 7.628468e+07 8.434376e+07 80818763 2018-12-31\n", "22 2019-06-01 9.425433e+07 9.040677e+07 9.820845e+07 92063064 2018-12-31\n", "23 2019-07-01 8.644319e+07 8.209196e+07 9.028962e+07 76972694 2018-12-31\n", "24 2019-05-01 8.133520e+07 7.723318e+07 8.567094e+07 80818763 2019-04-02\n", "25 2019-06-01 9.609183e+07 9.203950e+07 1.001646e+08 92063064 2019-04-02\n", "26 2019-07-01 8.867596e+07 8.443326e+07 9.269204e+07 76972694 2019-04-02\n", "27 2019-08-01 8.127349e+07 7.731661e+07 8.532102e+07 79035290 2019-04-02\n", "28 2019-09-01 8.582029e+07 8.163789e+07 8.968897e+07 91128679 2019-04-02\n", "29 2019-10-01 1.034187e+08 9.929004e+07 1.076682e+08 100686490 2019-04-02\n", "30 2019-08-01 8.131430e+07 7.717230e+07 8.549134e+07 79035290 2019-07-02\n", "31 2019-09-01 8.525117e+07 8.106940e+07 8.974195e+07 91128679 2019-07-02\n", "32 2019-10-01 1.020143e+08 9.796074e+07 1.062698e+08 100686490 2019-07-02\n", "33 2019-11-01 9.165255e+07 8.748693e+07 9.568568e+07 79899088 2019-07-02\n", "34 2019-12-01 9.346174e+07 8.930227e+07 9.781523e+07 103536412 2019-07-02\n", "35 2020-01-01 1.055244e+08 1.011898e+08 1.097860e+08 91491919 2019-07-02" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Define initial training size, horizon and period\n", "days_per_month = 30.436875\n", "CV_INITIAL_SIZE = 24*days_per_month\n", "CV_PERIOD_SIZE = 3*days_per_month\n", "CV_HORIZON_SIZE = 6*days_per_month\n", "\n", "# Run cross validation\n", "df_cv = cross_validation(m, initial = pd.to_timedelta(CV_INITIAL_SIZE, unit='D'),\n", " period = pd.to_timedelta(CV_PERIOD_SIZE, unit='D'),\n", " horizon = pd.to_timedelta(CV_HORIZON_SIZE, unit='D'))\n", "df_cv['cutoff'] = pd.to_datetime(df_cv['cutoff'].dt.date)\n", "df_cv" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In the above DataFrame, we can see all the predicted values per cutoff point along with the uncertainty intervals. To illustrate this in a more clear manner, we will plot the actual and predicted values for all the simulated historical forecasts." ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [], "source": [ "def plot_predictions(m, df_cv):\n", " \"\"\" \n", " Plots actual and predicted values for all different cutoff points \n", " Input args:\n", " m: Prophet fitted model\n", " df_cv: Output DataFrame of Prophet cross_validation method\n", " \"\"\"\n", " i=1\n", " # Iterate through the cutoff points\n", " for ct in df_cv['cutoff'].unique():\n", " dfs = df_cv.loc[df_cv['cutoff'] == ct]\n", " fig = plt.figure(facecolor='w', figsize=(8, 4))\n", " ax = fig.add_subplot(111)\n", " ax.plot(m.history['ds'].values, m.history['y'], 'k.')\n", " ax.plot(dfs['ds'].values, dfs['yhat'], ls = '-', c = '#0072B2')\n", " ax.fill_between(dfs['ds'].values, dfs['yhat_lower'], dfs['yhat_upper'], color = '#0072B2', alpha = 0.4)\n", " ax.axvline(x=pd.to_datetime(ct), c = 'gray', lw = 4, alpha = 0.5)\n", " ax.set_title('Cutoff date: {} \\n Training set size: {} months'.\\\n", " format(pd.to_datetime(ct).date(), int((CV_INITIAL_SIZE + i*CV_PERIOD_SIZE) / days_per_month)))\n", " ax.set_ylabel('Quantity')\n", " ax.set_xlabel('Date')\n", " i+=1\n", " \n", " \n", "def get_cv_performance_metrics(df_cv):\n", " \"\"\" \n", " Calculate basic performance metrics for the cross-validated results \n", " Input args:\n", " df_cv: Output DataFrame of Prophet cross_validation method\n", " \"\"\"\n", " dfp = performance_metrics(df_cv)\n", " dfpg = dfp.copy()\n", " dfpg['horizon_days'] = dfpg['horizon'].map(lambda x: round(x.days,-1))\n", " dfpg = dfpg.groupby('horizon_days').agg({'horizon': 'first',\n", " 'rmse': 'mean',\n", " 'mae': 'mean',\n", " 'mape': 'mean',\n", " 'coverage': 'mean'}).\\\n", " reset_index(drop=True)\n", " return dfp, dfpg" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfIAAAElCAYAAAAfqmBHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3dd3hUVf4/8PdkJnUy6YUkE0oIYAgQpKiRFkC/QBRU+NJ+FBEx0my7YldsX9FV0VUUxGURpO2CK0UhiNEoygCClIUIhBJIgTTSk8lMZs7vj5CRSZlMkqnJ+/U8PA8z98y958wlfHLuKR+JEEKAiIiInJKLvStARERErcdATkRE5MQYyImIiJwYAzkREZETYyAnIiJyYgzkREREToyBnMgOzp49i1tvvRUKhQIfffQRqqqqMH78ePj6+mLy5MnNfj41NRVKpdIGNSUiR8dATnTDpk2bMGjQIHh7eyMsLAzjxo3DL7/8YtZnu3btiu+//97sa/3tb39DQkICysrK8Pjjj2Pbtm3Izc1FYWEhtm7d2tomNOqLL77A0KFDLXa+b7/9FkOHDoWfnx86deqERx55BGVlZYbj1dXVmDt3Lnx8fNCpUycsX77c6PNJSUno1asXXFxc8MUXXxgdE0LgpZdeQkREBHx9fZGQkIDTp0+brM8HH3yATp06wdfXF3PnzkV1dbXhmLe3t9EfqVSKxx57rMlzXb9+HQ888ADkcjm6dOmCTZs2GY4dPHgQd999NwICAhAcHIzJkyfj6tWr5nxlRFbFQE4EYPny5XjyySfxwgsvIDc3F1euXMHChQuxY8cOq1zv8uXLiI2NNXrds2dPyGQyq1zPkkpKSvDSSy8hJycHf/zxB7KysrBkyRLD8VdffRXp6em4fPkyfvzxR/ztb39DcnKy4XhcXBw+/fRTDBgwoMG5t27din/+85/Yv38/rl+/jvj4eMyaNavJuuzduxdvv/02UlJSkJGRgYsXL2Lp0qWG4+Xl5YY/ubm58PT0NPnEY9GiRXBzc0Nubi42btyIBQsWGH6RKCoqQlJSEjIyMnD58mUoFAo89NBDLfruiKxCEHVwxcXFQi6Xi3//+99NlnnwwQfFiy++aHj9448/ioiICCGEEDNnzhQSiUR4eHgIuVwu3nnnHSGEEDt27BC9e/cWvr6+YsSIESItLU0IIcTIkSOFi4uLcHd3F3K5XEybNk24uroKmUwm5HK5+Mc//tHg+pWVleLBBx8Ufn5+IiYmRvztb38zXF8IIZYtWyaioqKEt7e3iImJEf/5z3+EEEKkpaUJd3d34eLiIuRyufD19RVCCKFWq8Vf//pXERkZKUJCQsSjjz4qKisrW/X9ffXVV6JPnz6G1+Hh4WLv3r2G1y+99JKYOnVqg88NGTJErF271ui9t99+W0yePNnw+tSpU8Ld3b3Ja0+fPl08//zzhtfff/+9CA0NbbTsF198Ibp16yb0en2jx8vLy4Wrq6s4e/as4b2ZM2eKZ599ttHyR48eFd7e3k3WjchW2COnDk+lUkGtVuOBBx5o1ee//PJLdO7cGbt27UJ5eTmeeeYZnDt3DtOnT8eHH36I/Px8JCYmYvz48dBoNPjhhx8wbNgwrFixAuXl5di8eTNeeOEFTJ06FeXl5Xj44YcbXOO1117DhQsXcOHCBezduxfr1q0zOt69e3fs378fJSUlWLp0KWbOnImrV68iJiYGq1atQnx8PMrLy1FcXAwAePbZZ3Hu3DkcP34c58+fR3Z2Nl5//XXD+fz8/MweVvj5558NTxeKioqQk5ODuLg4w/G4uLhmH4/XmTZtGs6fP49z585Bq9Vi3bp1GDt2bJPlT58+3eBadUMU9a1btw6zZ8+GRCJp9Fznzp2DVCpFz549zar7ze0msicGcurwCgsLERQUZNHH2v/6179wzz334O6774arqyuefvppVFVV4cCBA60637///W+8+OKLCAgIQGRkJB5//HGj45MnT0Z4eDhcXFwwdepU9OjRA4cPH270XEIIfP755/jggw8QEBAAhUKBF154AVu2bDGUKS4uNmtcfd++fVi3bp3hl4Dy8nIAgK+vr6GMr6+v0Ri6KWFhYRg2bBh69eoFT09PbN26FR988EGT5cvLyxtcC0CD6125cgU//fQTHnzwQbPPZaruJ0+exOuvv453333XrHYRWRMDOXV4gYGBKCgoQE1NjcXOmZOTgy5duhheu7i4IDIyEtnZ2a0+X2RkpOH1zecGgPXr16N///7w8/ODn58fTp06hYKCgkbPlZ+fj8rKSgwcONBQfuzYscjPz29RnQ4ePIj/9//+H7Zt22boxXp7ewMASktLDeVKS0uhUCjMOudrr72G3377DZmZmVCr1Vi6dClGjRqFyspKbNy40TBpbdy4cYbr1b8WgAbXW79+PYYOHYpu3boZ3hs3bpzhfHXnvvlcTdX9/PnzGDduHP7+979j2LBhZrWLyJoYyKnDi4+Ph4eHB7Zv395kGblcjsrKSsPra9euGR2v/7g2PDwcly9fNrwWQiAzMxMRERGtqmNYWBgyMzMNr69cuWL4++XLl/HII49gxYoVKCwsRHFxMfr06QNxI7Fh/boFBQXB09MTp0+fRnFxMYqLi1FSUmLoTZvj2LFjmDBhAv75z39i9OjRhvf9/f0RFhaGEydOGN47ceKE2Y+gT5w4galTp0KpVEImk2HOnDkoKipCWloaZsyYYZi4tmfPHgBAbGxsg2uFhoYiMDDQ6Lzr169v0Bvfs2eP4XwzZsxAz549UVNTg/T09CbrfvnyZdx11114+eWXTU7CI7IlBnLq8Hx9ffH6669j0aJF2L59OyorK6HVarFnzx4888wzAID+/ftj9+7duH79Oq5du4YPP/zQ6ByhoaG4ePGi4fWUKVPw7bffIiUlBVqtFu+//z7c3d1x5513tqqOU6ZMwbJly1BUVISsrCx8/PHHhmMVFRWQSCQIDg4GAKxduxanTp0yqltWVhY0Gg2A2qcDjzzyCJ566ink5eUBALKzs7F3716z6nLq1CmMHTsWH3/8McaPH9/g+OzZs/Hmm2+iqKgIZ86cweeff445c+YYjms0GqjVagghoNVqoVarodfrAQCDBw/G1q1bkZubC71ejy+//BJarRbR0dGN1mX27NlYs2YN0tLSUFRUhDfffNPoWgBw4MABZGdnN7s+Xy6XY+LEiXjllVdQUVGBX3/9FTt27DAE7OzsbIwaNQqLFi3C/PnzzfquiGzCvnPtiBzHhg0bxMCBA4WXl5cIDQ0ViYmJ4tdffxVCCFFVVSWmTJkiFAqF6Nu3r1i+fLnRrPHt27eLyMhI4evrK959910hhBD/+c9/RExMjPDx8RHDhw8Xp06dMpQfMWKE+Pzzzw2vly5dKmbMmNFk3SoqKsSsWbOEr69vo7PWX3jhBeHv7y8CAwPFU089JYYPH244f3V1tUhMTDQcr2vP888/L7p16yYUCoW45ZZbxN///nfD+eRyufj5558brcucOXOERCIRcrnc8Kd3796G42q1Wjz00ENCoVCIkJAQ8f777xt9fsSIEQKA0Z8ff/zRUK+FCxeKTp06CYVCIW699VaxZ8+eJr8XIYR4//33RUhIiFAoFGLOnDlCrVYbHU9KShIzZ840eY46hYWF4r777hNeXl4iMjJSbNy40XDs1VdfFQCM2i2Xy806L5E1SYS48fyNiIiInA4frRMRETkxBnIiIiInxkBORETkxBjIiYiInBgDOVELXLx40bDpiSXLOoNevXph//799q6GU3jppZcaLIMjshYGcmq3rly5YpTCUiKRQC6XG163JihFRUWZvXFKS8railKpRGpqaqs+e/bsWavuZKbX6zFy5EgEBwfD19cX/fv3xzfffGNUZsOGDejSpQu8vb0xceJEw97x9vT999+ja9eu9q4GdWAM5NRude7c2SiNJVC7U1fd68aCkk6ns3U16QaJRIKPPvoIV69eRUlJCT799FNMnz7dsGnNyZMnsXDhQmzcuBHXrl2Dq6srFi9ebOdaE9kfAzl1aDNnzsSiRYswduxYyOVy7N+/Hzt37kT//v2hUCjQuXNnvPHGG4by58+fN9rydOjQoVi6dCnuvPNOKBQKjB07FtevX29xWaB2R7bOnTsjKCgIb731lsne8zfffIOYmBgoFAoolUqjxCI7d+5EXFwc/Pz8MHToUMMub9OnT0dOTo5hj/Hly5c3OG9eXh4SExPh5+eHgIAADB8+3HDs5vrc/KRDLpdDIpEgKyvL5PWbI5FI0LdvX8hkMggh4OLiAo1GYzjvhg0bcP/992Po0KHw9vbG66+/jq1btxptnXszpVKJ9957D3369IG3tzeSkpKQm5uLMWPGwMfHB//zP/9j1KPfvn07YmNj4efnh1GjRuHs2bNG51q+fDn69u0LX19fTJ8+HdXV1SgpKcH48eONnv7U/eJRXV2NmTNnQqFQoE+fPvj9998N53vrrbcQHh4OHx8f3HLLLa1+SkIEgDu7UccBQKSnpxu9N2PGDOHn5ycOHDggdDqdUKvVIiUlRfz3v/8VOp1OHD9+XAQGBopdu3YJIYRIT08XN//YDBkyRERHR4tz586JiooKMXToUEPe8paUPXnypPD29ha//vqrUKvV4sknnxRSqdSw41l9QUFBhl3nCgsLxdGjR4UQQhw+fFiEhISIw4cPi5qaGrFmzRoRFRUlqqurhRBCRERENHlOIYR4+umnxaJFi4RGoxHV1dUiNTXVcKypzy5ZskQkJCQIrVbb7PWTkpLEY4891uT1hRBi7Nixws3NTQAQiYmJQqfTCSGESExMFO+9955RWQ8PD3H8+PFGzxMRESHi4+NFbm6uyMzMFAEBAWLgwIHi+PHjoqqqSgwfPly8+eabQojavO1yuVykpKQIjUYj/u///k/06NFDaDQaw7luv/12cfXqVVFQUCB69Ohh2Dlv3759okuXLkbXfvHFF4WHh4dITk4WNTU14umnnxZDhgwRQtTmWO/cubO4evWqEEKIixcvigsXLpj8TohMccoe+dy5cxESEoI+ffo0W/bKlSsYOXIkbr31VvTr1w+7d++2QQ3JmTzwwAOIj4+Hi4sL3N3dMWrUKPTp0wcuLi6Ii4vDtGnT8NNPPzX5+Ycffhg9evSAl5cXJk+ejOPHj7e47NatW3H//ffjzjvvhLu7O958802TdXZ1dUVaWhrKysoQEBCAAQMGAABWr16NhQsXYvDgwZBKpZg7dy4A4LfffjPru3B1dUVOTg6uXLkCNzc3jBgxwmT5jRs3Ytu2bdi6dStkMlmz1//ss8/w0UcfmTxnXTKTb775BnfffTdcXGr/m2oszaiPj4/JFKlPPPEEQkJCoFQqMXToUMTHxyMuLg4eHh64//77cezYMQDAli1bMGHCBIwaNQqurq547rnnUFpaikOHDhnO9eSTT6JTp04IDAzEvffea/I+A8CIESMwZswYSKVSzJo1y1BeJpNBrVbj9OnTqKmpQbdu3RAVFWXyXESmOGUgnzNnDpKTk80q++abb2LKlCk4duwYtmzZgoULF1q5duRsbk4PCgAqlQoJCQmGSVf/+Mc/mkwJCgCdOnUy/N3Ly8vkBLemytZPUyqXy+Hv79/keb7++mvs3LkTnTt3RkJCgiHgXL58Ge+8844hPamfnx+uXr1qdvrU5557Dl26dMHo0aPRvXt3k/m2jxw5gieffBLbt29HUFCQRa5fx9XVFffccw+++eYbwy/f5qYZvVloaKjh756eng1e3/z91087q1QqjerdkvvcWPmKigoAtbP/33//fbzyyisICQnB9OnTG2TTI2oJpwzkw4cPR0BAgNF7Fy5cwNixYzFw4EAMGzYMZ86cAVA77lb3w19SUoLw8HCb15ccW/00n9OmTcOkSZOQmZmJkpISzJs3z5AS1FrCwsIMY8FAbUazoqKiJsvffvvt2LlzJ/Ly8nDvvfdi2rRpAGp/KVm6dKkhPWlxcTEqKysxZcoUAA3bWp+Pjw8++OADZGRkYPv27XjnnXcafRqRm5uLiRMnYtWqVejXr5/h/eau31I1NTW4cOECgIYpS8+dOwe9Xo8ePXq06tw3q592Vq/XIysry6y0s819p42ZOXMmfv31V1y6dAk6nQ7PP/98i89BVMcpA3ljkpKS8PHHH+Po0aN47733DD3vV199FRs2bIBSqURiYqJR+keixtQ9rvbw8MDBgwexZcsWq19z8uTJ2L59Ow4ePAiNRoNXXnmlybJVVVXYtGkTSktL4erqCoVCAalUCqD25+CTTz7Bb7/9BiEEysvLsWvXLkNvsH661fp27dqFCxcuQAgBX19fSKVSw7nraLVaTJw4EQ899BAmTZpkdKy565uSlpaG5ORkqNVqaDQarFu3DgcOHDBMuJs5cya2b9+OAwcOoKKiAq+88gomT54MLy+vZs/dnClTpmDnzp1ITU2FVqvFu+++C4VCgdtvv73Zz4aGhqKgoMDkI/6b/fHHH/jxxx9RXV0NT09PeHp6NviOiVqiXQTy8vJyHDhwAJMnT0b//v3x6KOP4urVqwCAzZs3Y86cOcjKysLu3bsxa9YsQ+5josasXLkSzz//PBQKBd56661W9yZbol+/fvjggw8wefJkhIeHIzAwEIGBgXB3d2+0/Lp169ClSxf4+PhgzZo1+PLLLwHU9tRXrlyJBQsWwN/fHz179sSGDRsMn3vhhRewdOlS+Pn5NcipDtSuFR81ahS8vb0xZMgQPPHEExg6dKhRmcuXL+PAgQN4//33jWav5+TkNHv9efPmNblkTK/XGx43h4SE4NNPP8W2bdsQFxdn+I5WrFiBadOmISQkBNXV1Rb7xTw2Nhbr1q3DggULEBwcjOTkZOzcuROurq7NfrZPnz6YNGkSunbtCj8/P8Os9aZUV1fjmWeeQVBQEDp16mTIo07UWk6bxjQjIwP33nsvTp06hdLSUvTq1csQvG8WGxuL5ORkw/hjVFQUDh48iJCQEFtXmchspaWl8PPzw+XLlxuM4RMR3axd9Mh9fHzQrVs3bN26FQAghDCMpXXu3BkpKSkAah9pqdVqBAcH262uRE3ZuXMnKisrUV5ejr/+9a8YMGAAgzgRNcspA/n06dMRHx+Ps2fPQqlUYs2aNdi4cSPWrFmDuLg4xMbGYseOHQCA999/H59//jni4uIwffp0fPHFF62anEJkbV9//TXCw8OhVCqRkZGBzZs327tKROQEnPbROhERETlpj5yIiIhqyexdgZYKCgpipiEiMlv9ZWGmNpAhclQZGRlNbkzldIG8a9euOHLkiL2rQUROon5CkoSEBLvUg6gtBg0a1OQxPlonIiJyYgzkREREToyBnIiIyIkxkBMRETkxBnIiIiInxkBORETkxBjIiYiIzKRSqbBs2TKoVCp7V8XA6daRExER2YNKpcLo0aOh0Wjg5uaGlJQUxMfH27ta7JETERGZIzU1FRqNBjqdDhqNpsFmQ/bCQE5ERGSGhIQEuLm5QSqVws3NzWF2CeSjdSIiIjPEx8cjJSUFqampSEhIcIjH6gADORERkdni4+MdJoDXsdqj9blz5yIkJAR9+vRp9LgQAo8//jiio6PRr18//P7779aqChERdXCOONvcUqwWyOfMmYPk5OQmj+/Zswfp6elIT0/H6tWrsWDBAmtVhYiIOrC62eYvv/wyRo8e3e6CudUC+fDhwxEQENDk8R07dmD27NmQSCS44447UFxcjKtXr1qrOkRE1EE56mxzS7HbrPXs7GxERkYaXiuVSmRnZzdadvXq1Rg0aBAGDRqE/Px8W1WRiIjaAUedbW4pdpvsJoRo8J5EImm0bFJSEpKSkgCYTq5ORERUn6PONrcUuwVypVKJzMxMw+usrCyEh4fbqzpERNSOOeJsc0ux26P1CRMmYP369RBC4ODBg/D19UVYWJi9qkNEROSUrNYjnz59OlJTU1FQUAClUonXXnsNWq0WADB//nwkJiZi9+7diI6OhpeXF9auXWutqhAREbVbVgvkmzdvNnlcIpHgk08+sdbliYiIOgTutU5EROTEGMiJiIicGAM5ERGRE2MgJyIicmIM5ERERE6MgZyIiMiJMZATERE5MQZyIiIiJ8ZATkRE5MQYyImIyGpUKhWWLVsGlUpl76q0W3bLfkZERO2bSqXC6NGjodFo4ObmhpSUlHabgayOSqWyebpUBnIiIrKK1NRUaDQa6HQ6aDQapKamWi242SOANlYHe/ziwkBORERWkZCQADc3N0NgS0hIsMp1HKXnb8tfXG7GQE5ERFYRHx+PlJQUq/eU7RVA67PVLy71MZATEZHVxMfHWz2o2iuA1merX1zqYyAnIiKnZq8A2lRdbH19BnIiInJ69gigjoLryImILKSkSovskip7V4M6GAZyIiILuXS9EoeuFNm7GtTBMJATEVlIQYUG10qr7V0N6mAYyImILORaWTUKK7T2rgZ1MAzkREQWUlChQXl1DdRanb2rQh0IAzkRkQVUVNfgM9VlFFdpUVZdY+/qUAfCQE5EZAGHrhTjTF45Ll6vZCAnm2IgJyKygD/yygAAReyRk40xkBMRWUB6fgUAoFRdg4IKjZ1rQx0JAzkRkQVcKKwN5CVqLfLLuQSNbIeBnIjIAjKu1+7oVlylRV4ZAznZDgM5EZEF5JSqIQGg0QlklajtXR3qQBjIiYjaSFOjR1GlFiEKdwBAfrkGlRpOeCPbYCAnImqj9IIKCABd/T0B1I6Tl1U77qYwKpUKy5Ytg0qlsndVyAKYxpSIqI3SrtUuPesW4IVDV4pRrK5BWXUNQm/00B2JSqXC6NGjodFo4ObmhpSUlA6b/rO9sGqPPDk5Gb169UJ0dDTefvvtBsdLSkowfvx4xMXFITY2FmvXrrVmdYiIrOJMXjkAIFThDk9XFxRValGmdsxH66mpqdBoNNDpdNBoNEhNTW31udizdwxW65HrdDosWrQI+/btg1KpxODBgzFhwgT07t3bUOaTTz5B7969sWvXLuTn56NXr16YMWMG3NzcrFUtIiKLu1BYAZmLBF6uLgj0ckNxlRb5FY45cz0hIQFubm6GHnlCQkKrzsOeveOwWo/88OHDiI6ORlRUFNzc3DBt2jTs2LHDqIxEIkFZWRmEECgvL0dAQABkMj7tJyLncqGwEr4eMsjdZAj38UCJugZ55Y65KUx8fDxSUlLwxhtvtCn4WrJn7yic9QmD1aJmdnY2IiMjDa+VSiUOHTpkVGbx4sWYMGECwsPDUVZWhn/9619wceH8OyJyLpnFVfD3dEWQtxu6BnjiaFYx8hx4U5j4+Pg2954t1bN3FM78hMFqUVMI0eA9iURi9Hrv3r3o378/cnJycPz4cSxevBilpaUNPrd69WoMGjQIgwYNQn5+vrWqTETUYkII5JZVw8/TFaHe7ugeKIdOAJlFVY3+P9heWKpnbw5b9JSd+QmD1XrkSqUSmZmZhtdZWVkIDw83KrN27Vo899xzkEgkiI6ORrdu3XDmzBncdtttRuWSkpKQlJQEABg0aJC1qkxE1GKFFRqoa/Tw9XRFqMId0cFyAMD1Si0qNDp4u7ff4UJL9OybY6uesjM/YbBaj3zw4MFIT0/HpUuXoNFosGXLFkyYMMGoTOfOnZGSkgIAyM3NxdmzZxEVFWWtKhERWdzp3NqlZ/6ervDzdMUtwd4A6taSW2fmui16qI4yXmyrnrItnzBYmtV+VZTJZFixYgXGjBkDnU6HuXPnIjY2FqtWrQIAzJ8/Hy+//DLmzJmDvn37QgiBd955B0FBQdaqEhGRxaXl1i49C/ByhY+HDJ1urB0vrqpdSx5m4evZoofqSOPFtuwp2+IJgzVY9ZlPYmIiEhMTjd6bP3++4e/h4eH47rvvrFkFIiKrqktf6ushg8JdBh8PV3i7SXG9SoNStdbi12ush2rp4GOLa9RRqVRITU1FQkJCo9eo6ymbKtPRtd/BGyIiG7hQWAG5mxTuMim83Wr/Sw32dkNJlRb5VliCZoseqq16web2/J21p2wrDORERG2Qcb0Sfh6uCPByhYtL7cqccB8PnLpWZpW15LboodqqF2zLnn97xkBORNQG2SVqdFK4I9j7z33VuwV44UBGEXKtlJfcFj1UW1zDmWeKOxIGciKiVtLq9LheqcUtId4I9f5za+moIDkEajeK0euFoadOxjj+bRkM5ERErXShsDZ9qb+nq1GPvEdQ7Vry4ioNyjU18PFwtVMNHR/Hv9uO+6ESEbXS6RvpS/293ODj8We/KCa0di15sboG5Q6cl5zaBwZyIqJWOnsjfam/pww+N+3gVrcpTHGV1ipL0IhuxkBORNRK6QWVkEokkLtJobipRy53l8HXQ4brlVqUtDAvuV4vDD19InMwkBMRtdLFwgr4etamL3WXSY2OhXi730hn2rKZ6wUVGuy/WGjJalI7x0BORNRKmcVV8LuRvrS+CF8PlKq1yK9oWSDPLa9GaQt78dSxMZATEbXStbJq+HnUpi+tLyrQC2XVOlwrUbfonBcKKqDV6S1VReoAGMiJiFrhekU1qrR6+HvKEKpoGMi7B9YuQcssqYZOb35e8vOFFRarI3UMDORERK1Ql/XM38sNfp4N14n3uJGXvCXpTMura/Dzheu4WFhpuYpSu8cNYYiIWsEQyD1djdaQ14kJuZGXvKoG5dU1jQb7+vLKq/HTxUJ09vO0bGWpXWOPnIioFc4V1AbyuvSl9fUM9oaLBCiq0pg9ee1YVglK1TUWCeQqlQrLli2DSqVq87nIsbFHTkTUCufzK+DlKoWH65/pS2/mJnOBn6criiprUGzmpjDfp+cDADr7e7SpbuamB6X2gT1yIqJWyLheBT9PmVH60vo6KdxRotYiz4wsaDq9wLHsUrhKJQhTtC2QN5YelNovBnIiolbILlXDr16ylPqUvp4oUdcg34y85Pnl1cgsrkIXP09I25gtrS49qFQqZXrQDoCBnIiohWrTl2purCFvuBlMnahAL1RqdcgpbX4t+ZWiKlwrq0a3QK82168uPegbb7zBx+odAMfIiYha6FJhJfQC8Pcy3SOPvpHONLtEDa1OD1dp032n787lQy+AqIC2B3KA6R/rVuEAAB+1SURBVEE7EvbIiYha6FRubVKTAK/Gl57V6XljLXlZdU2za8kPZFwHAIT5uDe65StRUxjIiYha6OyNNeR+nq5G6UvrL/mqW0teXKVFuYlAXqauwYWCSoR6u6FG/2caVCJz8NE6EVELnS+sgFQCyF3/TF/a2JKv22+/HTIXCa5XaU2uJb9WpkZWiRpx4T4AACU3hKEWYI+ciKiFLhZWwtfTFd7uf6YvbWzJl4uLCwK8XFFUqUVRVdNryQ9fKUKlVofugV64knYMG1Z+yI1cyGwM5ERELXSlqAp+HsbpS5ta8mVYS24iL/n36Tfyj+dfwuYX5+KN15Zi9OjRDOZkFgZyIqIWulZWDT9P4/SlTS356uzvidIba8kb2za1RqfHyZwSeLq6IOfkr9BptdzIhVrErDHySZMmYe7cuRg3bhxcXBj7iajjKqrUoFKrg18j6UsbW/LVPVCOb2rycPDQIfzlxYcabJuaX6FBVokaUQFyKIMHw83NDVqthhu5kNnMisoLFizApk2b0KNHDzz33HM4c+aMtetFROSQ0m4sPWsqfWl9dXnJT5861ei2qen55cgr16B7oBciY/pj++5kbuRCLWJWj/yuu+7CXXfdhZKSEmzevBl33303IiMj8cgjj2DmzJlwdW3+HzMRkTWoVCqkpqYiISHBJoHvjxtLzwKaSF9aX68ba8l9ovrC1c0N0Bj3tvedKwAAdPH3hIerFP+TMAxjRg63TuWpXTJ7+VlhYSE2bNiAL7/8ErfeeitmzJiBX375BevWreM4DhHZhT2yfJ3LN52+tL6YTgoAgDQoEhu//hZnfz9o9EuHKuM6JKjdJS46yAsSSdv2WaeOx6xAPnHiRJw5cwazZs3Crl27EBYWBgCYOnUqBg0aZNUKEhE1pbElX9YO5BcKK+Hp6tJk+tL6lL4ecJNKcL1Si+i+AzFxzEjDsVK1FhevV0Lp54EavUB0IDeCoZYzK5DPmzcPiYmJRu9VV1fD3d0dR44csUrFiIiaU7fkS6Ox3eSwS4WV8PN0NZm+9GYSiQRBcjcUV2lRWGm8BC2nRI3sEjXiuwYAkCDct23pS6ljMmuy20svvdTgPXN+601OTkavXr0QHR2Nt99+u9Eyqamp6N+/P2JjYzFixAhzqkNEBMA+Wb6yS9Xw83BFsNz8/dDDFB4oUdcgr8w4nen+S4XQ6AS6+XtC6gKEcI91agWTPfJr164hOzsbVVVVOHbsGIQQAIDS0lJUVlaaPLFOp8OiRYuwb98+KJVKDB48GBMmTEDv3r0NZYqLi7Fw4UIkJyejc+fOyMvLs0CTiKgjsWWWrxqdHgUVGkQHyhssPTOlS4AnTl4rRX69TWF+PF+7EUwnH3dE+nlCZiI7GlFTTAbyvXv34osvvkBWVhb+8pe/GN5XKBR46623TJ748OHDiI6ORlRUFABg2rRp2LFjh1Eg37RpEyZOnIjOnTsDAEJCQlrdECIia8u4Xpu+1K+Z9KX1dQ+UQ6sTyCz+My+5VqfHqatl8PGQQeYiwS0hHB+n1jEZyB988EE8+OCD+OqrrzBp0qQWnTg7OxuRkZGG10qlEocOHTIqc+7cOWi1WiQkJKCsrAxPPPEEZs+e3aLrEBHZSloLl57VqctLnl+hgVqrg4erFPnltRvBdA/0ggCg9GWiFGodk/8SN2zYgJkzZyIjIwPLly9vcPzmXnp9dY/hb1Z/WUVNTQ2OHj2KlJQUVFVVIT4+HnfccQd69uxpVG716tVYvXo1ACA/P99UlYmIrOaPvNrNYOqnL21Or5DaQF6irs2C5uEqxelrpSiq0mJUQCAgJAhVcHycWsfkv8SKigoAQHl5eYNjza11VCqVyMzMNLzOyspCeHh4gzJBQUGQy+WQy+UYPnw4Tpw40SCQJyUlISkpCQC43I2I7EKlUmFb8n8hQSd4u/+ZvtQcvUNq15KXVNWgXFODELjju7O1nZIIXw8EebvBy4ylbESNMfkv59FHHwVQu7PbkCFDjI79+uuvJk88ePBgpKen49KlS4iIiMCWLVuwadMmozL33XcfFi9ejJqaGmg0Ghw6dAhPPfVUa9pBRGQ1dRvPVI1aDIQCheer4H5P7+Y/eEOwwh2eri4orNQY8pIfulIMmYsEPh4yw+5vRK1h1hTJxx57zKz3biaTybBixQqMGTMGMTExmDJlCmJjY7Fq1SqsWrUKABATE4OxY8eiX79+uO222zBv3jz06dOnFc0gIrKeuo1n4BsKlFxD/pmjLT5HsGEtuQYlVVpcul6JLv6e0Auga4CXFWpNHYXJHrlKpcKBAweQn59vNEZeWloKnU7X7MkTExMbbCQzf/58o9dLlizBkiVLWlJnIiKbqtt4psq3EyTnD+L2+/+nxecI9/XAubwK5JZWI7O4ClfLqjEqOhAS1OYsJ2otkz1yjUaD8vJy1NTUoKyszPDHx8cH27Zts1UdiYjaJDMzE/v37zeat9MS8fHx+OrbvYCnD269cwSG1xtqNEdXfy+UVNcgv7waP5wvgE4v0MXPC15uMrOyqBE1xWSPfMSIERgxYgTmzJmDLl262KpOREQWk5mZifXr10On00EqlSI6OrpVG8hIw3sBOIRuXbvA36vlgbd7oBw6vcCV4ipcLqoCAATKmSiF2s6saZLV1dVISkpCRkYGampqDO//8MMPVqsYEZElZGRkQKfTQQgBnU7X6sQqv2eVAAD8W7iGvE6PGxPaCis1OJ1bhiC5G2QuLugRxIlu1DZm/WucPHky5s+fj3nz5kEqlVq7TkREFtO1a1dIpVJDj7y1iVWO59QGcl9P89KX1tfrxs5tJeoaZBar0aeTApAIhPkwUQq1jVn/GmUyGRYsWGDtuhARWVxkZCRmz56NjIwMdO3atdHeuEqlQmpqqlGe8JvV6PQ4X1AJL1cpPF1lZqUvrS82tHYtecb1KlRodOgW4AXpjcxoRG1h1r/G8ePH49NPP8UDDzwAd/c/Z1cGBARYrWJERJYSGRlptGX0zerWiNelQm0si1peuQY5JWqE+bibnb60PoWHDN5uUpzJq91gK1Thji7+XkyUQm1mViBft24dAODdd981vCeRSHDx4kXr1IqIyEbq1ojrdDpoNJpGx9DP5pXhWlk1RkYHtih9aX0h3u64eL0S7jIXyN1c0JMbwZAFmBXIL126ZO16EBHZRd0a8boeeWNj6N/+kQcBIMLXs0XpS+uL8PXAxeuViArwAiCB0o+JUqjtzB7oOXXqFNLS0qBW/5mGj5nKiMjZxcfHIyUlpckxcp1e4ODlIkglEnRSuLUofWl93QK8sP/SdUQF1u7k1pZfCojqmBXIX3vtNaSmpiItLQ2JiYnYs2cPhg4dykBORO1CfHx8k0vS8surkVFUia4BnnCXSVu19KxO9xtLzcJ9PBCqcIenK1cBUduZNcti27ZtSElJQadOnbB27VqcOHEC1dXV1q4bEXVwKpUKy5Ytg0qlslsd0vMrcLW02rDeuyXpS+u7LzYUt3f2Q4jCjYlSyGLM+hfp6ekJFxcXyGQylJaWIiQkhBPdiMiqzJlNbgvf/pELvcCfgdyj9dupRgXKcVfPIEBI0MWfiVLIMszqkQ8aNAjFxcV45JFHMHDgQAwYMAC33XabtetGRO2cqR53Y7PJbU2vFzhwuQguEiBY4YbOfp5wk7V+uZjcTQoJJNBDoBM3giELMatH/umnnwKozVw2duxYlJaWol+/flatGBG1b831uM2ZTW5t+RUaZBRWorOfJzQ1esSEerfpfC4uEgR4uUKt1TNRClmMWYH8559/bvS94cOHW7xCRNQxNLd+u7nZ5LZwsbAcOaVqjO4RDCGAzv5tXy4WJHfjJDeyKLMC+c0bwajVahw+fBgDBw5k0hQiajVzetymZpPbwrdpedAJGDKUdVK0/XG40tcTQd7clpUsx6xAvmvXLqPXmZmZeOaZZ6xSISLqGByhx22KXi/wS0YRJKhd791J4dGm8fE6w7sHgklLyZJatY5CqVTi1KlTlq4LEXUw9u5xm1JQocGlwgpE+nlCq9OjdxvHx+tIW7FPO5EpZgXyxx57zJD4Xq/X49ixY4iLi7NqxYiI7OnS9Upkl1YjoXugxcbHiazBrEB+yy23QKfTAQACAwMxffp0DBkyxKoVIyKypz1/5EKnF+gRJLfY+DiRNZgM5FqtFkuWLMH69evRtWtXCCGQl5eHxx57DEOGDMGxY8dw66232qquREQ2odcL7L90HRIAnRTuCPOxzPg4kTWYDOR//etfUVlZicuXL0OhUAAASktL8fTTT2PBggVITk5mZjQiapRKpXLYiWzNKazU4OL1SkT4ekCnFxYbHyeyBpOBfPfu3UhPTzeMjwOAj48PVq5ciaCgIOzZs8fqFSQi5+Mo26u21pWiKmQVqzE8KgB6IRDJdKPkwEw+K3JxcTEK4nWkUimCg4Nxxx13WK1iROS8HGF71bZIPpOHGr1A9I3x8TBup0oOzGQg7927N9avX9/g/Q0bNiAmJsZqlSIi51a32YtUKrXb9qqtJYRA6oVCAECYjzsi27i/OpG1mXy0/sknn2DixIn45z//iYEDB0IikeC3335DVVUVvv76a1vVkYicjKNv9mJKYYUGFwsrEe7jDr0eHB8nh2cykEdERODQoUP44YcfcPr0aQghMG7cOIwePdpW9SOimzjTBDJH3uzFlMwSNbJKqhDflePj5BzMWkc+atQojBo1ytp1ISITnH0CmbPYeyYPGp1Ajxv7q3N8nBwdB36InISzTyBzBkII/FQ3Pq7w4Pg4OQX+CyVyEs48gcxZFFVpcb6gAqEKdwigzfnHiWyhVUlTiMj2nHkCmbPILKpCZrEat3X2g4BAZ46PkxNgICdyIs46gcxZ7EvPR7VOX7t+HBJ08nG3d5WImmXVR+vJycno1asXoqOj8fbbbzdZ7rfffoNUKsW2bdusWR0isiCVSoVly5ZBpVLZuyoWk3q+dnw8wscdSj8PuMukdq4RUfOs1iPX6XRYtGgR9u3bB6VSicGDB2PChAno3bt3g3LPPvssxowZY62qEJGFtccZ9EWVGpzLr0Cw3A2QSNA7VGHvKhGZxWo98sOHDyM6OhpRUVFwc3PDtGnTsGPHjgblPv74Y0yaNAkhISHWqgoRWVh7nEGfXaJGZnEVegbLIcD14+Q8rBbIs7OzERkZaXitVCqRnZ3doMzXX3+N+fPnW6saRGQF7XEG/Y/phVDX/Dk+HsbxcXISVnu0LoRo8F79BCxPPvkk3nnnHUilpsehVq9ejdWrVwMA8vPzLVdJImqV9jiD/vfsYgBAhI8HInw5Pk7Ow2qBXKlUIjMz0/A6KysL4eHhRmWOHDmCadOmAQAKCgqwe/duyGQy3H///UblkpKSkJSUBAAYNGiQtapMRC3Q3mbQp+WVw89TBhcXCWI7cXycnIfVAvngwYORnp6OS5cuISIiAlu2bMGmTZuMyly6dMnw9zlz5uDee+9tEMSJyLKcab92WxFC4I/cckQHeXF8nJyO1QK5TCbDihUrMGbMGOh0OsydOxexsbFYtWoVAHBcnMgO2uNsc0tIyy1HWXUNuvp7ARwfJydj1Q1hEhMTkZiYaPReUwH8iy++sGZViAiNzzZnIIdhf/UQb3coOT5OToZ7rRN1IO1xtrkl3NbZD5P6dYJEIhATwv3Vyblwi1aiDqQ9zja3hEGRfpgSF47UC4XoEuBl7+oQtQgDOVEH095mm1uSj4crx8fJ6fDROhE10B73UTdHV39Pjo+T02GPnIiMdNSZ7S4SCcfHySmxR05ERtrjPurmCPF2R3Sw3N7VIGox9siJyEjdzPa6HnlHmdk+vHugvatA1CoM5ERkhDPbiZwLAzkRNcCZ7UTOg2PkREREToyBnIiIyIkxkBMRETkxBnIiIiInxkBO5EA66o5qRNR6nLVO5CA66o5qRNQ27JETOYiOuqMaEbUNAzmRg2CucCJqDT5aJ3IQ3FGNiFqDgZzIgVhiRzWVSsVfBog6EAZyonaEE+aIOh6OkRO1I5wwR9TxMJATtSOcMEfU8fDROlE7wglzRB0PAzlRO8MUpEQdCx+tExEROTEGciIiIifGQE5EROTEGMiJiIicGAM5ERGRE2Mgpw6POcCJyJlx+Rl1aNzSlIicHXvk1K4119u21Jam7NUTkb1YNZAnJyejV69eiI6Oxttvv93g+MaNG9GvXz/069cPd955J06cOGHN6lAHU9fbfvnllzF69OhGg6wltjQ15zpERNZitUCu0+mwaNEi7NmzB2lpadi8eTPS0tKMynTr1g0//fQTTp48iZdffhlJSUnWqg51QOb0tuu2NH3jjTda/VidiUqIyJ6sNkZ++PBhREdHIyoqCgAwbdo07NixA7179zaUufPOOw1/v+OOO5CVlWWt6lAHVNfbrhv/bqq33dYtTc29DhGRNVgtkGdnZyMyMtLwWqlU4tChQ02WX7NmDcaNG2et6pATUqlUJpN/NHfcVglEmKiEiOzJaoFcCNHgPYlE0mjZH3/8EWvWrMEvv/zS6PHVq1dj9erVAID8/HzLVZIcVnOzyc2dbW6rBCJMVEJE9mK1MXKlUonMzEzD66ysLISHhzcod/LkScybNw87duxAYGBgo+dKSkrCkSNHcOTIEQQHB1uryuRAmht35rg0EVEtqwXywYMHIz09HZcuXYJGo8GWLVswYcIEozJXrlzBxIkT8eWXX6Jnz57Wqgo5oeZmk1titjkRUXtgtUfrMpkMK1aswJgxY6DT6TB37lzExsZi1apVAID58+fj9ddfR2FhIRYuXGj4zJEjR6xVJXIizY07O+O4dHNj+kRErSERjQ1mO7BBgwYx2JPT4Q5y9lN/2IVPb8gZmYp93NmNyAY4pk9E1sJATmQDHNMnImth0hQiG3DGMX0icg4M5EQ2wrXmRGQNfLRORETkxBjIOxim22w5fmdE5Mj4aN0GHGX9sKMtgXKU78UUR/vOiIjqYyC3MkcKBI0tgbJXXRzpezHFkb4zIqLG8NG6lTnS+mFHWgLlSN+LKY70nRERNYY9citzpFzVjrQEypG+F1Mc6TsjImoMt2i1AWcYC7YHfi9kC9yildoDU7GPPfJmWCLYcP1w4/i9EBG1HQO5Cc4yIYuIiDouTnYzwdwJWY6yzthR6kFERLbDHrkJ5kzIcpReu6PUg4iIbIs9chPqZiy/8cYbTQZGR1lG5Sj1qMOnA0REtsEeeTOam5DlKMuoHKUeAJ8OEBHZEgN5G1lqnXFbZ8c70npn7oZGRGQ7DOQW0NZlVOb0YM0J9JZYzmWJ5XaO9HSAiKi969CB3FE2JGmuB2urR9XmXqe5782Rng4QEbV3HTaQO9I4bnM9WFs9qjbnOuZ+b9zshYjINjrsrHVHmuXd3Ox4WyXuMOc6jvS9ERFRB+6RO9o4rqkerK0eVZtzHUf73oiIOroOnTTFUcbInQ2/N3ImTJpC7QGTpjSB47itw++NiMhxdNgxciIiovaAgZyIiMiJMZATERE5MQZyIiIiJ8ZATkRE5MQYyImIiJyY060jDwoKQteuXc0un5+fj+DgYOtVyIbaU1vqtKc2tae2AGyPo2tP7WlPbQGs056MjAwUFBQ0eszpAnlLWXIDGXtrT22p057a1J7aArA9jq49tac9tQWwfXv4aJ2IiMiJMZATERE5Memrr776qr0rYW0DBw60dxUspj21pU57alN7agvA9ji69tSe9tQWwLbtafdj5ERERO0ZH60TERE5MQZyIiIiJ+ZwgTwzMxMjR45ETEwMYmNj8fe//x0AcP36ddx9993o0aMH7r77bhQVFQEACgsLMXLkSHh7e2Px4sVG59JoNEhKSkLPnj1xyy234Kuvvmr0mkePHkXfvn0RHR2Nxx9/HHWjDT///DMGDBgAmUyGbdu2OXVbnnrqKfTv3x/9+/dHz5494efnZ7f2lJWVGerSv39/BAUF4cknn2xRe9p6bxytPY50fwBg8+bN6Nu3L/r164exY8c2uX7VGX522toWS9wbS7fpX//6F/r164fY2Fg888wzTV7TGe5PW9tij5+dffv2YeDAgejbty8GDhyIH374odl6mtueVt0b4WBycnLE0aNHhRBClJaWih49eojTp0+LJUuWiGXLlgkhhFi2bJl45plnhBBClJeXi/3794uVK1eKRYsWGZ3rlVdeES+++KIQQgidTify8/MbvebgwYPFgQMHhF6vF2PHjhW7d+8WQghx6dIlceLECTFr1iyxdetWp27LzT766CPx0EMP2bU9NxswYID46aefWtSett4bR2vPzex9f7RarQgODjb8G1uyZIlYunRpi9rjKD87lmjLzVp7byzZpoKCAhEZGSny8vKEEELMnj1bfP/99y1qk6PcH0u05Wa2+tn5/fffRXZ2thBCiP/+978iPDy8RfU0Va4198bhAnl9EyZMEN99953o2bOnyMnJEULUfuk9e/Y0Krd27doG/7kqlUpRXl5u8vw5OTmiV69ehtebNm0SSUlJRmUefPDBVgeLmzlCW4QQIj4+Xnz33XetbYZBW9pT59y5c0KpVAq9Xt/gmC3vjRCO0R4h7H9/NBqNCAoKEhkZGUKv14tHH31UfPbZZ61qj71/dizZFiEsd2/a0qbDhw+L0aNHG16vX79eLFiwoMH5neH+WLItQtj+Z0cIIfR6vQgICBBqtdrselr63jjco/WbZWRk4NixY7j99tuRm5uLsLAwAEBYWBjy8vJMfra4uBgA8PLLL2PAgAGYPHkycnNzG5TLzs6GUqk0vFYqlcjOzrZgK2o5SlsuX76MS5cuYdSoUXZrz802b96MqVOnQiKRNDhmq3sDOE57HOH+uLq6YuXKlejbty/Cw8ORlpaGhx9+uFXtsQRHaYul7k1b2xQdHY0zZ84gIyMDNTU12L59OzIzM1vVJktwlLbY62fnq6++wq233gp3d3ezv3NL3xuHDeTl5eWYNGkSPvzwQ/j4+LT48zU1NcjKysKQIUPw+++/Iz4+Hk8//XSDcqKR8YvG/hNuC0dqy5YtW/C///u/kEqlLa5Hnba2p359pk+f3ugxW9wbwLHa4wj3R6vVYuXKlTh27BhycnLQr18/LFu2rEE5Z/jZsWRbLHFvgLa3yd/fHytXrsTUqVMxbNgwdO3aFTKZrEE5Z7g/lmyLPX52Tp8+jWeffRafffaZ2fVsSTlzOWQg12q1mDRpEmbMmIGJEycCAEJDQ3H16lUAwNWrVxESEmLyHIGBgfDy8sIDDzwAAJg8eTJ+//136HQ6w8SIV155BUqlEllZWYbPZWVlITw8vN22xVSgsVV76pw4cQI1NTWGjRNsfW8csT2OcH+OHz8OAOjevTskEgmmTJmCAwcOOOXPjiXb0tZ7Y6k2AcD48eNx6NAhqFQq9OrVCz169HDK+2PJttj6ZycrKwsPPPAA1q9fj+7duwNAk/W09r1xuEAuhMDDDz+MmJgY/OUvfzG8P2HCBKxbtw4AsG7dOtx3330mzyORSDB+/HikpqYCAFJSUtC7d29IpVIcP34cx48fx+uvv46wsDAoFAocPHgQQgisX7++2XM7a1vOnj2LoqIixMfH27U9dTZv3mz0g2fLe+OI7XGU+xMREYG0tDTk5+cDqJ2hGxMT45Q/O5ZqS1vvjSXbBMDwiLeoqAiffvop5s2b55T3x1JtsfXPTnFxMe655x4sW7YMQ4YMMZRvqp5WvzdmjaTb0P79+wUA0bdvXxEXFyfi4uLEt99+KwoKCsSoUaNEdHS0GDVqlCgsLDR8pkuXLsLf31/I5XIREREhTp8+LYQQIiMjQwwbNkz07dtXjBo1Sly+fLnRa/72228iNjZWREVFiUWLFhkmKh0+fFhEREQILy8vERAQIHr37u20bRFCiKVLl4pnn322RW2wVnuEEKJbt27ijz/+MHlNa90bR2uPEI51f1auXCluueUW0bdvX3HvvfeKgoKCFrXHkX522toWIdp+byzdpmnTpomYmBgRExMjNm/e3OQ1neH+tLUtQtj+Z+eNN94QXl5ehrJxcXEiNze32Xqa057W3Btu0UpEROTEHO7ROhEREZmPgZyIiMiJMZATERE5MQZyIiIiJ8ZATkRE5MQYyIk6MKlUiv79+yM2NhZxcXFYvnw59Hq9yc9kZGRg06ZNNqohETWHgZyoA/P09MTx48dx+vRp7Nu3D7t378Zrr71m8jMM5ESOhYGciAAAISEhWL16NVasWAEhBDIyMjBs2DAMGDAAAwYMwIEDBwAAzz33HPbv34/+/fvjgw8+gE6nw5IlSzB48GD069fPsO80EdkGN4Qh6sC8vb1RXl5u9J6/vz/OnDkDhUIBFxcXeHh4ID09HdOnT8eRI0eQmpqK9957D9988w0AYPXq1cjLy8NLL72E6upqDBkyBFu3bkW3bt3s0SSiDqdhmhki6tDqfrfXarVYvHgxjh8/DqlUinPnzjVa/rvvvsPJkyexbds2AEBJSQnS09MZyIlshIGciAwuXrwIqVSKkJAQvPbaawgNDcWJEyeg1+vh4eHR6GeEEPj4448xZswYG9eWiACOkRPRDfn5+Zg/fz4WL14MiUSCkpIShIWFwcXFBV9++SV0Oh0AQKFQoKyszPC5MWPGYOXKldBqtQCAc+fOoaKiwi5tIOqI2CMn6sCqqqrQv39/aLVayGQyzJo1y5DGceHChZg0aRK2bt2KkSNHQi6XAwD69esHmUyGuLg4zJkzB0888QQyMjIwYMAACCEQHByM7du327NZRB0KJ7sRERE5MT5aJyIicmIM5ERERE6MgZyIiMiJMZATERE5MQZyIiIiJ8ZATkRE5MQYyImIiJzY/wcUOUAgahqbfAAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_predictions(m, df_cv)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, let's extract some useful statistics for the prediction performance. We will look specifically on the performance of the model for various horizon sizes." ] }, { "cell_type": "code", "execution_count": 23, "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", "
horizonrmsemaemapecoverage
030 days1.014294e+077.943737e+060.1328220.222222
160 days9.612280e+067.865447e+060.1134350.166667
290 days1.215714e+079.925678e+060.1443490.138889
3121 days1.184559e+071.033440e+070.1583600.222222
4151 days7.461642e+066.733523e+060.0896860.138889
5182 days9.629865e+068.470354e+060.1095120.444444
\n", "
" ], "text/plain": [ " horizon rmse mae mape coverage\n", "0 30 days 1.014294e+07 7.943737e+06 0.132822 0.222222\n", "1 60 days 9.612280e+06 7.865447e+06 0.113435 0.166667\n", "2 90 days 1.215714e+07 9.925678e+06 0.144349 0.138889\n", "3 121 days 1.184559e+07 1.033440e+07 0.158360 0.222222\n", "4 151 days 7.461642e+06 6.733523e+06 0.089686 0.138889\n", "5 182 days 9.629865e+06 8.470354e+06 0.109512 0.444444" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dfp, dfpg = get_cv_performance_metrics(df_cv)\n", "display(dfpg)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 1.0, 'Cross-Validated MAPE for various horizon sizes')" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Plot MAPE with respect to horizon\n", "plt.figure(figsize = (6,4))\n", "plt.plot(dfpg['mape'], '*-')\n", "labels = [str(h.days) + ' days' for h in dfpg['horizon']]\n", "plt.xticks(np.arange(len(dfpg)), labels = labels, rotation = 20)\n", "plt.grid()\n", "plt.xlabel('Horizon')\n", "plt.ylabel('MAPE')\n", "plt.title('Cross-Validated MAPE for various horizon sizes')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As can be seen from the above MAPE plot, using 3.5 years of Vortexa's US Crude export data and a basic forecasting model with default settings and no parametrization we were able to predict exports in a 6 month horizon with MAPE up to 8%. Not bad at all! " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Make Future Predictions" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "After the cross-validation we can now move to the final and most exciting part. Finally, let's predict how much crude oil United States will export in the upcoming 6 months." ] }, { "cell_type": "code", "execution_count": 25, "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", "
ds
482020-02-01
492020-03-01
502020-04-01
512020-05-01
522020-06-01
532020-07-01
\n", "
" ], "text/plain": [ " ds\n", "48 2020-02-01\n", "49 2020-03-01\n", "50 2020-04-01\n", "51 2020-05-01\n", "52 2020-06-01\n", "53 2020-07-01" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Create a DataFrame with future dates\n", "MONTHS_AHEAD_TO_PREDICT = 6\n", "future = m.make_future_dataframe(periods = MONTHS_AHEAD_TO_PREDICT, freq = 'MS')\n", "future.tail(MONTHS_AHEAD_TO_PREDICT)" ] }, { "cell_type": "code", "execution_count": 26, "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", "
dsyhatyhat_loweryhat_upper
482020-02-011.029069e+089.815953e+071.073748e+08
492020-03-011.088316e+081.040407e+081.135400e+08
502020-04-011.113489e+081.067104e+081.160252e+08
512020-05-011.148086e+081.098006e+081.197779e+08
522020-06-011.081688e+081.035143e+081.131841e+08
532020-07-011.140947e+081.094229e+081.191747e+08
\n", "
" ], "text/plain": [ " ds yhat yhat_lower yhat_upper\n", "48 2020-02-01 1.029069e+08 9.815953e+07 1.073748e+08\n", "49 2020-03-01 1.088316e+08 1.040407e+08 1.135400e+08\n", "50 2020-04-01 1.113489e+08 1.067104e+08 1.160252e+08\n", "51 2020-05-01 1.148086e+08 1.098006e+08 1.197779e+08\n", "52 2020-06-01 1.081688e+08 1.035143e+08 1.131841e+08\n", "53 2020-07-01 1.140947e+08 1.094229e+08 1.191747e+08" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Make predictions\n", "forecast = m.predict(future)\n", "forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail(MONTHS_AHEAD_TO_PREDICT)" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 33.0, 'Date')" ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Plot forecasts along with historical data\n", "fig = m.plot(forecast)\n", "plt.plot(forecast.set_index('ds')['yhat'].tail(MONTHS_AHEAD_TO_PREDICT), '*', color = 'black', label = 'Forecasted values')\n", "plt.legend(loc = 'upper left')\n", "plt.title('US Crude Exports - {} months Forecasting'.format(MONTHS_AHEAD_TO_PREDICT))\n", "plt.ylabel('Quantity (barrels)')\n", "plt.xlabel('Date')" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [], "source": [ "#fig2 = m.plot_components(forecast)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Conclusion" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For a baseline model with practically no fine tuning at all, the results look promising. Those data combined with additional regressors from the SDK (e.g. diesel exports from Rotterdam or crude exports from Russia) or even with proprietary data can create even more powerful models and shed light to the future patterns and trends of the global oil seaborne flows. Hoped you enjoyed!" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "vortexa_sdk", "language": "python", "name": "vortexa_sdk" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.6" } }, "nbformat": 4, "nbformat_minor": 2 }