{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "df = pd.read_parquet(\"fraud.parquet\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train/test split\n",
    "\n",
    "We're using time-series data, so we'll split based on time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "first = df['timestamp'].min()\n",
    "last = df['timestamp'].max()\n",
    "cutoff = first + ((last - first) * 0.7)\n",
    "\n",
    "df = df.sample(frac=0.1).copy()\n",
    "\n",
    "train = df[df['timestamp'] <= cutoff]\n",
    "test = df[df['timestamp'] > cutoff]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cloudpickle as cp\n",
    "feature_pipeline = cp.load(open('feature_pipeline.sav', 'rb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Weighting samples\n",
    "\n",
    "We're going to weight samples by the the inverse of the frequency of their label."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "fraud_frequency = train[train[\"label\"] == \"fraud\"][\"timestamp\"].count() / train[\"timestamp\"].count()\n",
    "train[\"weights\"] = fraud_frequency\n",
    "train.loc[train[\"label\"] == \"fraud\", \"weights\"] = (1 - fraud_frequency)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# WIP from here..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'feature_pipeline' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-12-a3b16769fd33>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mlr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mLogisticRegression\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmax_iter\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m500\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0msvecs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfeature_pipeline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msmol_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      5\u001b[0m \u001b[0mlr\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msvecs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msmol_train\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"label\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msmol_train\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"weights\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'feature_pipeline' is not defined"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn import model_selection\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=4, max_depth=3, random_state=404, class_weight=\"balanced_subsample\")\n",
    "\n",
    "svecs = feature_pipeline.fit_transform(train)\n",
    "rfc.fit(svecs, train[\"label\"], sample_weight=train[\"weights\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "predictions = rfc.predict(feature_pipeline.fit_transform(test))\n",
    "print(classification_report(smol_test.label.values, predictions))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlworkflows import plot\n",
    "df, chart = plot.binary_confusion_matrix(test[\"label\"], predictions)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# save model here"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}