Skip to content

Instantly share code, notes, and snippets.

@inigoreiriz

inigoreiriz/prediction_intervals.ipynb

Created April 16, 2018 18:21
Show Gist options
  • Save inigoreiriz/2946551fcb521e4fecd0db15c9de65a4 to your computer and use it in GitHub Desktop.
Save inigoreiriz/2946551fcb521e4fecd0db15c9de65a4 to your computer and use it in GitHub Desktop.
Download ZIP
prediction_intervals
Display the source blob
Display the rendered blob
Raw
prediction_intervals.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python2.7/dist-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
" from ._conv import register_converters as _register_converters\n",
"Using TensorFlow backend.\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2018-04-16 14:23:47 - Removed 23737 outliers (3.53%) from train\n",
"2018-04-16 14:23:47 - Removed 8510 outliers (5.07%) from test\n",
"training DL model for mean\n",
"training DL model for quantile: 0.025\n",
"training DL model for quantile: 0.1\n",
"training DL model for quantile: 0.9\n",
"training DL model for quantile: 0.975\n",
"making predictions for quantile: 0.025\n",
"making predictions for quantile: 0.1\n",
"making predictions for quantile: 0.9\n",
"making predictions for quantile: 0.975\n",
"training DL model for multiple quantiles\n",
"Running quantile regression for quantile: 0.025\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running quantile regression for quantile: 0.1\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running quantile regression for quantile: 0.9\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running quantile regression for quantile: 0.975\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n",
"/usr/local/lib/python2.7/dist-packages/statsmodels/regression/quantile_regression.py:193: IterationLimitWarning: Maximum number of iterations (1000) reached.\n",
" \") reached.\", IterationLimitWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"MEAN ERRORS (ALL LINKS TOGETHER):\n",
"\tMAE\tRMSE\tR2\n",
"HA: 3.171 4.830 0.989\n",
"LR: 2.895 4.165 0.992\n",
"DL: 2.883 4.166 0.992\n",
"MULTI: 2.886 4.162 0.992\n",
"\n",
"\n",
"MEAN ERRORS (INDIVIDUAL LINKS):\n",
"\tMAE\tRMSE\tR2\n",
"HA: 0.210 0.278 0.240\n",
"LR: 0.200 0.262 0.307\n",
"DL: 0.199 0.261 0.303\n",
"MULTI: 0.197 0.260 0.311\n",
"\n",
"\n",
"PREDICTION INTERVAL: 95.0 %\n",
"\tICP\tMIL\tRMIL\n",
"QR: 0.937 0.971 38.311\n",
"DLQR: 0.943 0.929 37.853\n",
"MULTI: 0.894 0.781 41.730\n",
"HOMO: 0.028 0.017 0.677\n",
"\n",
"\n",
"PREDICTION INTERVAL: 80.0 %\n",
"\tICP\tMIL\tRMIL\n",
"QR: 0.774 0.598 23.612\n",
"DLQR: 0.767 0.580 22.378\n",
"MULTI: 0.778 0.532 28.125\n",
"HOMO: 0.018 0.011 0.443\n",
"\n",
"\n",
"[2-10] Crossing Cases: QR: 3\tDL: 0\t MULTI: 1\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 6\tDL: 9\t MULTI: 162\n",
"[2-10] Crossing Cases: QR: 0\tDL: 2\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 7\t MULTI: 0\n",
"[2-10] Crossing Cases: QR: 0\tDL: 13\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 10\t MULTI: 3\n",
"[2-10] Crossing Cases: QR: 0\tDL: 3\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 0\t MULTI: 30\n",
"[2-10] Crossing Cases: QR: 0\tDL: 5\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 1\t MULTI: 21\n",
"[2-10] Crossing Cases: QR: 0\tDL: 5\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 7\t MULTI: 8\n",
"[2-10] Crossing Cases: QR: 0\tDL: 13\t MULTI: 11\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 3\t MULTI: 14\n",
"[2-10] Crossing Cases: QR: 0\tDL: 3\t MULTI: 1\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 0\t MULTI: 64\n",
"[2-10] Crossing Cases: QR: 0\tDL: 1\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 9\t MULTI: 4\n",
"[2-10] Crossing Cases: QR: 0\tDL: 7\t MULTI: 1\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 5\t MULTI: 35\n",
"[2-10] Crossing Cases: QR: 0\tDL: 6\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 6\t MULTI: 0\n",
"[2-10] Crossing Cases: QR: 0\tDL: 3\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 2\t MULTI: 0\n",
"[2-10] Crossing Cases: QR: 0\tDL: 12\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 10\t MULTI: 6\n",
"[2-10] Crossing Cases: QR: 0\tDL: 12\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 6\t MULTI: 0\n",
"[2-10] Crossing Cases: QR: 0\tDL: 11\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 11\t MULTI: 2\n",
"[2-10] Crossing Cases: QR: 27\tDL: 2\t MULTI: 6\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 8\tDL: 1\t MULTI: 84\n",
"[2-10] Crossing Cases: QR: 1\tDL: 5\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 2\tDL: 3\t MULTI: 46\n",
"[2-10] Crossing Cases: QR: 0\tDL: 13\t MULTI: 4\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 14\t MULTI: 164\n",
"[2-10] Crossing Cases: QR: 0\tDL: 31\t MULTI: 3\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 8\t MULTI: 13\n",
"[2-10] Crossing Cases: QR: 6\tDL: 17\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 3\t MULTI: 122\n",
"[2-10] Crossing Cases: QR: 0\tDL: 3\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 35\t MULTI: 46\n",
"[2-10] Crossing Cases: QR: 0\tDL: 5\t MULTI: 3\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 4\n",
"[90-97] Crossing Cases: QR: 0\tDL: 40\t MULTI: 7\n",
"[2-10] Crossing Cases: QR: 0\tDL: 0\t MULTI: 8\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 2\t MULTI: 263\n",
"[2-10] Crossing Cases: QR: 6\tDL: 17\t MULTI: 4\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 4\n",
"[90-97] Crossing Cases: QR: 0\tDL: 8\t MULTI: 10\n",
"[2-10] Crossing Cases: QR: 0\tDL: 9\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 12\t MULTI: 71\n",
"[2-10] Crossing Cases: QR: 0\tDL: 2\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 1\tDL: 1\t MULTI: 43\n",
"[2-10] Crossing Cases: QR: 2\tDL: 13\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 1\tDL: 5\t MULTI: 16\n",
"[2-10] Crossing Cases: QR: 0\tDL: 5\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 20\t MULTI: 0\n",
"[2-10] Crossing Cases: QR: 0\tDL: 12\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 1\tDL: 16\t MULTI: 1\n",
"[2-10] Crossing Cases: QR: 0\tDL: 7\t MULTI: 0\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 7\t MULTI: 24\n",
"[2-10] Crossing Cases: QR: 0\tDL: 0\t MULTI: 133\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 0\n",
"[90-97] Crossing Cases: QR: 0\tDL: 1\t MULTI: 142\n",
"[2-10] Crossing Cases: QR: 0\tDL: 1\t MULTI: 19\n",
"[10-90] Crossing Cases: QR: 0\tDL: 0\t MULTI: 5\n",
"[90-97] Crossing Cases: QR: 0\tDL: 3\t MULTI: 68\n",
"\n",
"\n",
"Crossing Cases:\n",
"QR: 64\n",
"DL: 503\n",
"MULTI: 1676\n",
"\n",
"\n",
"Crossing Ratios:\n",
"QR: 0.000237\n",
"DL: 0.00186\n",
"MULTI: 0.00621\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAABS4AAALHCAYAAACAOtwWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQ3AJAAAAAAAAIOb8D3Fmduf1GjVHAAAAAElFTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7fadd2e54950>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA/gAAAHfCAYAAADk7XMwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7fadd2e49150>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAABUAAAAJFCAYAAAAcbF0GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7fadd2f79d50>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA/gAAAHfCAYAAADk7XMwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTAAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7fadd2f0bb10>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import time\n",
"import datetime\n",
"\n",
"import math\n",
"import numpy as np\n",
"import pandas as pd\n",
"import tensorflow as tf\n",
"\n",
"from sklearn import linear_model\n",
"import statsmodels.formula.api as smf\n",
"\n",
"from __future__ import division\n",
"\n",
"import keras\n",
"from keras.models import Sequential\n",
"from keras.layers import *\n",
"from keras.layers.wrappers import *\n",
"from keras.optimizers import RMSprop\n",
"from keras.callbacks import CSVLogger, EarlyStopping\n",
"import keras.backend.tensorflow_backend as ktf\n",
"from keras.callbacks import ModelCheckpoint\n",
"\n",
"from common2 import *\n",
"\n",
"import matplotlib.pyplot as plt\n",
"%matplotlib inline \n",
"from matplotlib.backends.backend_pdf import PdfPages\n",
"plt.style.use('ggplot')\n",
"\n",
"\n",
"def get_session(gpu_fraction=1):\n",
" gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=gpu_fraction,\n",
" allow_growth=True)\n",
" return tf.Session(config=tf.ConfigProto(gpu_options=gpu_options))\n",
"\n",
"ktf.set_session(get_session())\n",
"\n",
"def info(msg):\n",
" print(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S') + \" \" + msg)\n",
"\n",
"\"\"\" EVALUATION FUNCTIONS \"\"\"\n",
"\n",
"def compute_mean_errors(trues, preds):\n",
" \"\"\" Computes COR, MAE, MSE, RAE, RMSE, R2\n",
" args:\n",
" y_true: true values\n",
" y_pred: predicted values\n",
" \"\"\"\n",
" \n",
" trues = trues / 60\n",
" preds = preds / 60\n",
" \n",
" #pearson linear correlation coefficient\n",
" corr = np.corrcoef(preds, trues)[0,1]\n",
" #mean absolute error\n",
" mae = np.mean(np.abs(preds - trues))\n",
" #root mean square error\n",
" mse = np.mean((preds - trues)**2)\n",
" #root absolute error\n",
" rae = np.sum(np.abs(preds - trues)) / np.sum(np.abs(trues - np.mean(trues)))\n",
" #root mean square error \n",
" rmse = np.sqrt(np.mean((preds - trues)**2))\n",
" r2 = max(0, 1 - np.sum((trues-preds)**2) / np.sum((trues - np.mean(trues))**2))\n",
" return corr, mae, mse, rae, rmse, r2\n",
"\n",
"def compute_quantile_metrics(lower, upper, trues, preds):\n",
" \"\"\" Specific to quantiles. Computes ICP, MIL, RMIL \n",
" args:\n",
" lower: low quantile prediction (i.e, 5%)\n",
" upper: upper quantile prediction (i.e, 95%)\n",
" y_true: true values\n",
" y_pred: predicted_values\n",
" \"\"\"\n",
" \n",
" trues = trues / 60\n",
" preds = preds / 60\n",
" lower = lower / 60\n",
" upper = upper / 60\n",
" \n",
" N = len(trues)\n",
" icp = 1.0*np.sum((trues>lower) & (trues<upper)) / N\n",
" diffs = np.maximum(0, upper-lower)\n",
" mil = np.sum(diffs) / N\n",
" rmil = 0.0\n",
" for i in xrange(N):\n",
" if trues[i] != preds[i]:\n",
" rmil += diffs[i] / (np.abs(trues[i]-preds[i]))\n",
" rmil = rmil / N\n",
" return icp, mil, rmil\n",
"\n",
"\"\"\" LOSS FUNCTIONS \"\"\"\n",
"\n",
"def mean_squared_error(y_true, y_pred):\n",
" \"\"\" Default loss function in Keras \"\"\"\n",
" return K.mean(K.square(y_pred - y_true), axis=-1)\n",
"\n",
"def quantile_loss(q, y_true, y_pred):\n",
" \"\"\" Used for predicting one quantile for each model \n",
" args:\n",
" q: quantile\n",
" y_true: true values\n",
" y_pred: predicted values\n",
" \"\"\"\n",
" e = (y_true-y_pred) # y_true[:,:,:,:,0]-y_pred[:,:,:,:,0]\n",
" return K.mean(K.maximum(q*e, (q-1)*e), axis=-1)\n",
" \n",
"def multi_quantile_loss(quantiles, y_true, y_pred):\n",
" \"\"\" Model predicts multiple quantiles at a time\n",
" args:\n",
" quantiles: list of quantiles\n",
" y_true: true values\n",
" y_pred: predicted values\n",
" \"\"\"\n",
" #Last dimension contains mean + num_quantiles outputs\n",
" loss = K.mean(K.square(y_true[:,:,:,:,0]-y_pred[:,:,:,:,0]), axis=-1)\n",
" for k in xrange(len(quantiles)):\n",
" q = quantiles[k]\n",
" #extract quantile prediction\n",
" e = (y_true[:,:,:,:,k+1]-y_pred[:,:,:,:,k+1])\n",
" #compute quantile regr loss\n",
" loss += K.mean(K.maximum(q*e, (q-1)*e), axis=-1)\n",
" return loss\n",
"\n",
"def remove_outliers(data, low, upr): \n",
" _low = low.lookup(data['DowTimeRef'], data['LinkRef'])\n",
" _upr = upr.lookup(data['DowTimeRef'], data['LinkRef'])\n",
" mask = ((_low < data['LinkTravelTime']) & (data['LinkTravelTime'] < _upr))\n",
" data = data.loc[mask].copy()\n",
" return data, (~mask).sum()\n",
"\n",
"\"\"\" GLOBAL VARIABLES \"\"\"\n",
"\n",
"early_stopping = EarlyStopping(monitor='val_loss', patience=3)\n",
"\n",
"Ts = [1.282, 1.96]\n",
"quantiles = [0.025,0.1,0.9,0.975]\n",
"# number of timesteps (features) for LSTM\n",
"num_lags = 4 * 7\n",
"# number of predicitions ahead\n",
"num_preds = 1\n",
"# bootstrap_size (initial training size)\n",
"bootstrap_size = 0.80\n",
"# test window size\n",
"test_window = 0.20\n",
"# number of training epochs\n",
"num_epochs = 50\n",
"\n",
"LR_predictions_mean = []\n",
"LR_predictions_multi_mean = []\n",
"QR_predictions = []\n",
"DL_predictions_mean = []\n",
"DL_predictions_q = []\n",
"DL_predictions_multi_q = []\n",
"\n",
"hist_DL_mean = []\n",
"hist_DL_q = []\n",
"hist_DL_multi_q = []\n",
"\n",
"icps_links_qr, mils_links_qr, rmils_links_qr = ([] for i in range(3))\n",
"icps_links_dl, mils_links_dl, rmils_links_dl = ([] for i in range(3))\n",
"icps_links_multi, mils_links_multi, rmils_links_multi = ([] for i in range(3))\n",
"icps_links_homo, mils_links_homo, rmils_links_homo = ([] for i in range(3))\n",
"\n",
"\"\"\" PREPARE LINEAR QUANTILE REGRESSION MODEL \"\"\"\n",
"\n",
"col_names = [\"l%d\" % (i,) for i in xrange(num_lags)]\n",
"col_names.append(\"y\")\n",
"# 'y ~ l0+l1+l2+l3+l4...+l32'\n",
"qr_str = \"y ~ \" + ''.join([\"+l%d\" % (x,) for x in range(num_lags)])[1:]\n",
"\n",
"\"\"\" MODEL \"\"\"\n",
"\n",
"def build_simple_model(input_timesteps, output_timesteps, num_links, num_outs, loss=\"mse\"):\n",
" \n",
" model = Sequential()\n",
"\n",
" model.add(ConvLSTM2D(filters=20, kernel_size=(3, 3),\n",
" input_shape=(input_timesteps, num_links, 1, 1),\n",
" padding='same', return_sequences=True))\n",
" model.add(Dropout(0.2))\n",
"\n",
" model.add(ConvLSTM2D(filters=20, kernel_size=(3, 3),\n",
" padding='same', return_sequences=False))\n",
" \n",
"\n",
" model.add(Dropout(0.2))\n",
" \n",
" model.add(Flatten())\n",
" # Repeat last hidden state 'output_timesteps' times\n",
" model.add(RepeatVector(output_timesteps))\n",
" model.add(Reshape((output_timesteps, num_links, 1, 20)))\n",
" \n",
" model.add(Dense(units=num_outs))\n",
" model.add(Activation(\"linear\"))\n",
" \n",
" model.compile(loss=loss, optimizer=\"rmsprop\")\n",
" \n",
" return model \n",
" \n",
"\n",
"def build_model(input_timesteps, output_timesteps, num_links, num_outs, loss='mse'):\n",
"\n",
" \"\"\" DL MODEL \n",
" ENCODER: ConvLSTM2D -> ConvLSTM2D\n",
" DECODER: ConvLSTM2D -> ConvLSTM2D\n",
" args:\n",
" input_timesteps: number of lags/features\n",
" output_timesteps: number of timesteps we are predicting ahead\n",
" num_links: number of links we are training/predicting simultaneously\n",
" \"\"\"\n",
" # Notes: No need to provide batch_size as 1st dim in 'input_shape'\n",
" # fit() method needs 'num_samples' dim though\n",
" \n",
" model = Sequential()\n",
" \n",
" ### ENCODER (training) ### \n",
" \n",
" # input_shape(num_lags, num_links, XGRID, YGRID)\n",
" model.add(BatchNormalization(name = 'batch_norm_0', \n",
" input_shape = (input_timesteps, num_links, 1, 1)))\n",
" model.add(ConvLSTM2D(name ='conv_lstm_1',\n",
" filters = 64, kernel_size = (10, 1), \n",
" padding = 'same', \n",
" return_sequences = True)) \n",
"\n",
" # Notes:\n",
" ## return_sequences returns hidden states for each time step. Used when LSTM layer follows.\n",
" ## if return_sequences=False, time_step dim is eliminated. Thus, output cannot be fed to \n",
" ## an RNN.\n",
" \n",
" model.add(Dropout(0.2, name = 'dropout_1'))\n",
" model.add(BatchNormalization(name = 'batch_norm_1'))\n",
"\n",
" model.add(ConvLSTM2D(name ='conv_lstm_2',\n",
" filters = 64, kernel_size = (5, 1), \n",
" padding='same',\n",
" return_sequences = False))\n",
" # Notes:\n",
" ## No need to return_sequences now, since we we dont have an LSTM layer following.\n",
" ## Only return last hidden state output\n",
" \n",
" model.add(Dropout(0.1, name = 'dropout_2'))\n",
" model.add(BatchNormalization(name = 'batch_norm_2'))\n",
" \n",
" model.add(Flatten())\n",
" # Repeat last hidden state 'output_timesteps' times\n",
" model.add(RepeatVector(output_timesteps))\n",
" # Reshape. Input to decoder (output_timesteps, num_links, 1, 64)\n",
" model.add(Reshape((output_timesteps, num_links, 1, 64)))\n",
" \n",
" #### DECODER (INFERENCE) #### \n",
" \n",
" model.add(ConvLSTM2D(name ='conv_lstm_3',\n",
" filters = 64, kernel_size = (10, 1), \n",
" padding='same',\n",
" return_sequences = True))\n",
" \n",
" model.add(Dropout(0.1, name = 'dropout_3'))\n",
" model.add(BatchNormalization(name = 'batch_norm_3'))\n",
" \n",
" model.add(ConvLSTM2D(name ='conv_lstm_4',\n",
" filters = 64, kernel_size = (5, 1), \n",
" padding='same',\n",
" return_sequences = True))\n",
" \n",
" # Notes:\n",
" # returns hidden states for each time step (output_timesteps)\n",
" \n",
" # Apply dense layer across every output over time\n",
" model.add(TimeDistributed(Dense(units=num_outs, name = 'dense_1', activation = 'linear')))\n",
"\n",
" optimizer = RMSprop() #lr=0.0001, rho=0.9, epsilon=1e-08, decay=0.9)\n",
" model.compile(loss = loss, optimizer = optimizer)\n",
" return model\n",
"\n",
" \n",
"\"\"\" LOAD DATA \"\"\"\n",
"\n",
"\n",
"def load_data(path):\n",
" \n",
" data = prep_data(path)\n",
" data = data[(1 <= data['LineDirectionLinkOrder']) & (data['LineDirectionLinkOrder'] <= 32)]\n",
" data = data[data['LinkTravelTime'].notnull()]\n",
" assert len(data['LinkRef'].unique()) == 32\n",
" \n",
" return data\n",
"\n",
"\n",
"def models_infer(data):\n",
" \n",
" nrows = len(data)\n",
"\n",
" #Divide set into training and testing\n",
" j = 0\n",
" data_train = data[:int((bootstrap_size + j * test_window) * nrows)]\n",
" data_test = data[int((bootstrap_size + j * test_window) * nrows): \\\n",
" int((bootstrap_size + (j + 1) * test_window) * nrows)]\n",
"\n",
" # Extract mean, std, low and upper founds from each Link Travel Time Series subset\n",
" (means, scales, low, upr) = fit_scale(data_train)\n",
"\n",
" # safety check...\n",
" assert means.shape == (4 * 24 * 7, 32)\n",
" assert len(scales) == 32\n",
" assert low.shape == (4 * 24 * 7, 32)\n",
" assert upr.shape == (4 * 24 * 7, 32)\n",
"\n",
" data_train_no = data_train\n",
" data_train_no, n_outliers = remove_outliers(data_train, low, upr) \n",
" info('- Removed {0} outliers ({1:.2f}%) from train'.format(n_outliers, 100.0 * n_outliers / len(data_train)))\n",
" \n",
" data_test_no = data_test\n",
" data_test_no, n_outliers = remove_outliers(data_test, low, upr)\n",
" info('- Removed {0} outliers ({1:.2f}%) from test'.format(n_outliers, 100.0 * n_outliers / len(data_test)))\n",
" \n",
" # Remove mean and std from each Link Travel Time Series subset\n",
" ix_train, ts_train, rm_mean_train, rm_scale_train, w_train, lns_train = \\\n",
" transform(data_train_no, means, scales)\n",
" ix_test, ts_test, rm_mean_test, rm_scale_test, w_test, lns_test = \\\n",
" transform(data_test_no, means, scales)\n",
" \n",
" # Construct X with lags as features (mean and std removed)\n",
" # Constructs Y with num_preds we want to make ahead (mean and std removed)\n",
" X_train, Y_train, Y_ix_train, Y_rm_mean_train, Y_scale_train, Y_w_train = \\\n",
" roll(ix_train, ts_train, rm_mean_train, rm_scale_train, w_train, num_lags, num_preds)\n",
" X_test, Y_test, Y_ix_test, Y_rm_mean_test, Y_scale_test, Y_w_test = \\\n",
" roll(ix_test, ts_test, rm_mean_test, rm_scale_test, w_test, num_lags, num_preds)\n",
"\n",
" # 'input_shape' inside model must be 4 dimensional (num_timesteps, num_links, XGRID, YGRID )\n",
" # fit() method needs 5 dimensional array (num_samples, num_timesteps, num_links, XGRID, YGRID)\n",
" # thus, we need to add 2 additional dims to tensors (XGRID, YGRID)\n",
" X_train = X_train[:,:,:,np.newaxis,np.newaxis] \n",
" Y_train = Y_train[:,:,:,np.newaxis,np.newaxis]\n",
" X_test = X_test[:,:,:,np.newaxis,np.newaxis]\n",
" Y_test = Y_test[:,:,:,np.newaxis,np.newaxis]\n",
" \n",
" # Build Y_train_K for \"Multi Quantile Model\"\n",
"\n",
" Y_train_K = Y_train[:,:,:,:,:]\n",
" Y_test_K = Y_test[:,:,:,:,:]\n",
" for k in xrange(len(quantiles)):\n",
" Y_train_K = np.concatenate((Y_train_K, Y_train[:,:,:,:,:]), axis=4)\n",
" Y_test_K = np.concatenate((Y_test_K, Y_test[:,:,:,:,:]), axis=4)\n",
" \n",
" \"\"\" DL MODEL FOR MEAN \"\"\"\n",
"\n",
" ## Save model parameters\n",
" checkpoint = ModelCheckpoint(\"ConvLSTM_mean.hdf5\", monitor='val_loss', \n",
" verbose=0, save_best_only=True, mode='min')\n",
"\n",
" ## build model\n",
" DL_model_mean = build_simple_model(num_lags, num_preds, len(lns_train), num_outs = 1, \n",
" loss=lambda y_true,y_pred: mean_squared_error(y_true, y_pred))\n",
"\n",
" print \"training DL model for mean\" \n",
"\n",
" ## train model\n",
" history = DL_model_mean.fit(X_train, Y_train,\n",
" batch_size = 128, epochs = num_epochs,\n",
" shuffle = False, \n",
" validation_data = (X_test, Y_test),\n",
" callbacks=[checkpoint],\n",
" verbose = 0) \n",
" \n",
" hist_DL_mean.append(history)\n",
" \n",
" ## inference\n",
"\n",
" ## Load weights\n",
" DL_model_mean.load_weights(\"ConvLSTM_mean.hdf5\")\n",
"\n",
" ## Make predictions\n",
" DL_predictions_mean.append(DL_model_mean.predict(X_test))\n",
"\n",
" \"\"\" INDEPENDENT DL MODELS FOR EACH QUANTILE \"\"\"\n",
"\n",
" DL_models_q = []\n",
" # Train independent DL models for each quantile\n",
" for q in quantiles:\n",
" print \"training DL model for quantile:\", q\n",
" \n",
" # Save model parameters at each quantile step\n",
" checkpoint = ModelCheckpoint('ConvLSTM_q_' + str(q) + '.hdf5', monitor='val_loss', \n",
" verbose=0, save_best_only=True, mode='min')\n",
" \n",
" # build model\n",
" DL_model_q = build_simple_model(num_lags, num_preds, len(lns_train), num_outs = 1, \n",
" loss=lambda y_true,y_pred: quantile_loss(q,y_true,y_pred))\n",
" \n",
" # train model\n",
" history = DL_model_q.fit(X_train, Y_train,\n",
" batch_size = 128, epochs = num_epochs,\n",
" shuffle = False, \n",
" validation_data = (X_test, Y_test),\n",
" callbacks=[checkpoint],\n",
" verbose = 0)\n",
" \n",
" hist_DL_q.append(history)\n",
" \n",
" DL_models_q.append(DL_model_q)\n",
"\n",
" # Load trained DL models and predict each quantile\n",
" for i in xrange(len(quantiles)):\n",
" q = quantiles[i]\n",
" print \"making predictions for quantile:\", q\n",
"\n",
" # load weights\n",
" DL_models_q[i].load_weights('ConvLSTM_q_' + \n",
" str(q) + '.hdf5')\n",
" \n",
" # make predictions\n",
" DL_predictions_q.append(\n",
" DL_models_q[i].predict(X_test))\n",
"\n",
"\n",
" \"\"\" DL MODEL FOR MEAN + MULTIPLE QUANTILES \"\"\"\n",
"\n",
" ## Save model parameters\n",
" checkpoint = ModelCheckpoint(\"ConvLSTM_multi_quantile.hdf5\", monitor='val_loss', \n",
" verbose=0, save_best_only=True, mode='min')\n",
" # build model\n",
" DL_model_multi_q = build_simple_model(num_lags, num_preds, len(lns_train), num_outs = 1 + len(quantiles), \n",
" loss=lambda y_true,y_pred: multi_quantile_loss(quantiles, y_true, y_pred))\n",
"\n",
" print \"training DL model for multiple quantiles\"\n",
" \n",
" # train model. Note that Y_train is different. We have stacked last dimension num_quantiles times\n",
" # in order to get a prediction for each of the quantiles.\n",
" history = DL_model_multi_q.fit(X_train, Y_train_K,\n",
" batch_size = 128, epochs = num_epochs,\n",
" shuffle = False, \n",
" validation_data = (X_test, Y_test_K),\n",
" callbacks=[checkpoint],\n",
" verbose = 0) \n",
" \n",
" hist_DL_multi_q.append(history)\n",
" \n",
" # inference\n",
"\n",
" ## Load weights\n",
" DL_model_multi_q.load_weights(\"ConvLSTM_multi_quantile.hdf5\")\n",
"\n",
" ## Make predictions\n",
" DL_predictions_multi_q.append(DL_model_multi_q.predict(X_test)) \n",
"\n",
" \"\"\" BASELINE: INDEPENDENT LINEAR REGRESSION (MEAN) \"\"\"\n",
" \n",
" lr_preds = np.zeros((X_test.shape[0], len(lns_train)))\n",
" \n",
" for i in xrange(len(lns_train)):\n",
" \n",
" regr_ind = linear_model.LinearRegression()\n",
" regr_ind.fit(X_train[:,:,i,0,0], Y_train[:,0,i,0,0])\n",
" lr_preds[:,i] = regr_ind.predict(X_test[:,:,i,0,0])\n",
" \n",
" LR_predictions_mean.append(lr_preds)\n",
"\n",
" \"\"\" QUANTILE LINEAR REGRESSION \"\"\"\n",
" \n",
" # For each of the quantiles\n",
" for k in xrange(len(quantiles)):\n",
" q = quantiles[k]\n",
" print \"Running quantile regression for quantile:\", q\n",
"\n",
" qr_preds = np.zeros((X_test.shape[0], len(lns_train)))\n",
" \n",
" # For each of the links\n",
" for i in xrange(len(lns_train)):\n",
" \n",
" # Quantile Linear Regression\n",
" data = pd.DataFrame(data = np.hstack([X_train[:,:,i,0,0],\n",
" Y_train[:,0,i,0,:]]), columns = col_names) \n",
" \n",
" data_test = pd.DataFrame(data = np.hstack([X_test[:,:,i,0,0], \n",
" Y_test[:,0,i,0,:]]), columns = col_names)\n",
" \n",
" model_qr = smf.quantreg(qr_str, data)\n",
" model_qr = model_qr.fit(q=q)\n",
" qr_preds[:,i] = model_qr.predict(data_test)\n",
"\n",
" QR_predictions.append(qr_preds)\n",
"\n",
" return len(lns_train), Y_test, Y_scale_test, Y_rm_mean_test, Y_w_test\n",
"\n",
"\n",
"def evaluate_quantiles(num_links, Y_test, Y_scale_test, Y_rm_mean_test):\n",
"\n",
" maes_ha, rmses_ha, r2s_ha = ([] for i in range(3))\n",
" maes_lr, rmses_lr, r2s_lr = ([] for i in range(3))\n",
" maes_dl, rmses_dl, r2s_dl = ([] for i in range(3))\n",
" maes_multi, rmses_multi, r2s_multi = ([] for i in range(3))\n",
" \n",
" for i in range(num_links):\n",
"\n",
" lower_preds_homo = []\n",
" upper_preds_homo = []\n",
" lower_preds_qr = []\n",
" upper_preds_qr = []\n",
" lower_preds_dl = []\n",
" upper_preds_dl = []\n",
" lower_preds_multi = []\n",
" upper_preds_multi = []\n",
" \n",
" trues = Y_test[:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" \n",
" mean_preds_lr = LR_predictions_mean[0][:,i] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" \n",
" # lower bound preds qr\n",
" lower_preds_qr = []\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" lower_preds_qr.append(QR_predictions[j][:,i] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
" \n",
" # upper bound preds qr\n",
" upper_preds_qr = []\n",
" for j in range(int(len(quantiles)/2),len(quantiles)):\n",
" upper_preds_qr.append(QR_predictions[j][:,i] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
"\n",
" #######################\n",
" \n",
" mean_preds_dl = DL_predictions_mean[0][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" \n",
" # lower_preds_dl\n",
" lower_preds_dl = []\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" lower_preds_dl.append(DL_predictions_q[j][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
" \n",
" # upper_preds_dl\n",
" upper_preds_dl = []\n",
" for j in range(int(len(quantiles)/2),len(quantiles)):\n",
" upper_preds_dl.append(DL_predictions_q[j][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
" \n",
" #print(\"DL lower & upper bounds\\n\")\n",
" #print(np.mean(lower_preds_dl[0]))\n",
" #print(np.mean(upper_preds_dl[-1]))\n",
" #print(\"\\n\")\n",
" ########################\n",
" \n",
" mean_preds_multi = DL_predictions_multi_q[0][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" \n",
" # lower_preds_multi\n",
" lower_preds_multi = []\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" # j+1 since the 1st index contains mean prediction\n",
" lower_preds_multi.append(DL_predictions_multi_q[0][:,0,i,0,j+1] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
" \n",
" # upper_preds_multi\n",
" upper_preds_multi = []\n",
" for j in range(int(len(quantiles)/2),len(quantiles)):\n",
" upper_preds_multi.append(DL_predictions_multi_q[0][:,0,i,0,j+1] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i])\n",
" \n",
" #print(\"Multi lower & upper bounds\\n\")\n",
" #print(np.mean(lower_preds_multi[0]))\n",
" #print(np.mean(upper_preds_multi[-1]))\n",
" #print(\"\\n\")\n",
" \n",
" \"\"\" Evaluate Mean Predictions \"\"\"\n",
"\n",
" corr, mae, mse, rae, rmse, r2 = compute_mean_errors(trues, Y_rm_mean_test[:,0,i])\n",
" maes_ha.append(mae)\n",
" rmses_ha.append(rmse)\n",
" r2s_ha.append(r2)\n",
" \n",
" corr, mae, mse, rae, rmse, r2 = compute_mean_errors(trues, mean_preds_lr)\n",
" maes_lr.append(mae)\n",
" rmses_lr.append(rmse)\n",
" r2s_lr.append(r2)\n",
"\n",
" corr, mae, mse, rae, rmse, r2 = compute_mean_errors(trues, mean_preds_dl)\n",
" maes_dl.append(mae)\n",
" rmses_dl.append(rmse)\n",
" r2s_dl.append(r2)\n",
" \n",
" corr, mae, mse, rae, rmse, r2 = compute_mean_errors(trues, mean_preds_multi)\n",
" maes_multi.append(mae)\n",
" rmses_multi.append(rmse)\n",
" r2s_multi.append(r2)\n",
" \n",
" \"\"\" Homoskedastic quantile regression\"\"\"\n",
" \n",
" # Z score for 80% and 95% confidence intervals \n",
" for T in Ts[::-1]:\n",
" lower_preds_homo.append(mean_preds_dl - (T * rmses_dl[i] * (np.sqrt(1 + (1/len(mean_preds_dl))))))\n",
" \n",
" for T in Ts:\n",
" upper_preds_homo.append(mean_preds_dl + (T * rmses_dl[i] * (np.sqrt(1 + (1/len(mean_preds_dl))))))\n",
" \n",
" #print(\"Homo lower & upper bounds\\n\")\n",
" #print(np.mean(lower_preds_homo[0]))\n",
" #print(np.mean(upper_preds_homo[-1]))\n",
" #print(\"\\n\")\n",
" \n",
" \"\"\" Evaluate Quantile Predictions \"\"\"\n",
"\n",
" icps_link_qr, mils_link_qr, rmils_link_qr = ([] for i in range(3))\n",
" icps_link_dl, mils_link_dl, rmils_link_dl = ([] for i in range(3))\n",
" icps_link_multi, mils_link_multi, rmils_link_multi = ([] for i in range(3))\n",
" icps_link_homo, mils_link_homo, rmils_link_homo = ([] for i in range(3))\n",
" \n",
" # evaluate quantile linear regression model\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" icp, mil, rmil = compute_quantile_metrics(lower_preds_qr[j], upper_preds_qr[-(j+1)], \n",
" trues, mean_preds_dl)\n",
" \n",
" icps_link_qr.append(icp)\n",
" mils_link_qr.append(mil)\n",
" rmils_link_qr.append(rmil)\n",
" \n",
" # evaluate quantile deep learning model\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" icp, mil, rmil = compute_quantile_metrics(lower_preds_dl[j], upper_preds_dl[-(j+1)], \n",
" trues, mean_preds_dl)\n",
" \n",
" icps_link_dl.append(icp)\n",
" mils_link_dl.append(mil)\n",
" rmils_link_dl.append(rmil)\n",
" \n",
" # evaluate multi-quantile deep learning model\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" icp, mil, rmil = compute_quantile_metrics(lower_preds_multi[j], upper_preds_multi[-(j+1)], \n",
" trues, mean_preds_dl)\n",
" \n",
" icps_link_multi.append(icp)\n",
" mils_link_multi.append(mil)\n",
" rmils_link_multi.append(rmil)\n",
" \n",
" # evalute homosketastic quantile regression model\n",
" for j in range(0,int(len(quantiles)/2)):\n",
" icp, mil, rmil = compute_quantile_metrics(lower_preds_homo[j], upper_preds_homo[-(j+1)],\n",
" trues, mean_preds_dl)\n",
" \n",
" icps_link_homo.append(icp)\n",
" mils_link_homo.append(mil)\n",
" rmils_link_homo.append(rmil)\n",
" \n",
" ##########\n",
" \n",
" icps_links_qr.append(icps_link_qr)\n",
" mils_links_qr.append(mils_link_qr)\n",
" rmils_links_qr.append(rmils_link_qr)\n",
"\n",
" icps_links_dl.append(icps_link_dl)\n",
" mils_links_dl.append(mils_link_dl)\n",
" rmils_links_dl.append(rmils_link_dl) \n",
"\n",
" icps_links_multi.append(icps_link_multi)\n",
" mils_links_multi.append(mils_link_multi)\n",
" rmils_links_multi.append(rmils_link_multi) \n",
" \n",
" icps_links_homo.append(icps_link_homo)\n",
" mils_links_homo.append(mils_link_homo)\n",
" rmils_links_homo.append(rmils_link_homo) \n",
" \n",
" print \"MEAN ERRORS (INDIVIDUAL LINKS):\"\n",
" print \"\\tMAE\\tRMSE\\tR2\"\n",
" print \"HA: %.3f %.3f %.3f\" % (np.mean(maes_ha),np.mean(rmses_ha),np.mean(r2s_ha)) \n",
" print \"LR: %.3f %.3f %.3f\" % (np.mean(maes_lr),np.mean(rmses_lr),np.mean(r2s_lr)) \n",
" print \"DL: %.3f %.3f %.3f\" % (np.mean(maes_dl),np.mean(rmses_dl),np.mean(r2s_dl)) \n",
" print \"MULTI: %.3f %.3f %.3f\" % (np.mean(maes_multi),np.mean(rmses_multi),np.mean(r2s_multi)) \n",
" print \"\\n\"\n",
" \n",
" all_icp_q = []\n",
" all_mil_q = []\n",
" all_rmil_q = []\n",
" for k in range(0,int(len(quantiles)/2)):\n",
" \n",
" # i.e, icps for specific quantile from all links\n",
" lq_icp_qr, lq_mil_qr, lq_rmil_qr = ([] for i in range(3))\n",
" lq_icp_dl, lq_mil_dl, lq_rmil_dl = ([] for i in range(3))\n",
" lq_icp_multi, lq_mil_multi, lq_rmil_multi = ([] for i in range(3))\n",
" lq_icp_homo, lq_mil_homo, lq_rmil_homo = ([] for i in range(3))\n",
" \n",
" for j in range(num_links):\n",
" \n",
" lq_icp_qr.append(icps_links_qr[j][k])\n",
" lq_icp_dl.append(icps_links_dl[j][k])\n",
" lq_icp_multi.append(icps_links_multi[j][k])\n",
" lq_icp_homo.append(icps_links_homo[j][k])\n",
" \n",
" lq_mil_qr.append(mils_links_qr[j][k])\n",
" lq_mil_dl.append(mils_links_dl[j][k])\n",
" lq_mil_multi.append(mils_links_multi[j][k])\n",
" lq_mil_homo.append(mils_links_homo[j][k])\n",
" \n",
" lq_rmil_qr.append(rmils_links_qr[j][k])\n",
" lq_rmil_dl.append(rmils_links_dl[j][k])\n",
" lq_rmil_multi.append(rmils_links_multi[j][k])\n",
" lq_rmil_homo.append(rmils_links_homo[j][k])\n",
" \n",
" \n",
" print \"PREDICTION INTERVAL: \" + str((quantiles[-(k+1)] - quantiles[k]) * 100) + \" %\" \n",
" print \"\\tICP\\tMIL\\tRMIL\"\n",
" print \"QR: %.3f %.3f %.3f\" % (np.mean(lq_icp_qr), np.mean(lq_mil_qr), np.mean(lq_rmil_qr)) \n",
" print \"DLQR: %.3f %.3f %.3f\" % (np.mean(lq_icp_dl), np.mean(lq_mil_dl), np.mean(lq_rmil_dl)) \n",
" print \"MULTI: %.3f %.3f %.3f\" % (np.mean(lq_icp_multi), np.mean(lq_mil_multi), np.mean(lq_rmil_multi)) \n",
" print \"HOMO: %.3f %.3f %.3f\" % (np.mean(lq_icp_homo), np.mean(lq_mil_homo), np.mean(lq_rmil_homo)) \n",
" print \"\\n\"\n",
" \n",
" all_icp_q.append((np.mean(lq_icp_qr), np.mean(lq_icp_dl), np.mean(lq_icp_multi)))\n",
" all_mil_q.append((np.mean(lq_mil_qr), np.mean(lq_mil_dl), np.mean(lq_mil_multi)))\n",
" all_rmil_q.append((np.mean(lq_rmil_qr), np.mean(lq_rmil_dl), np.mean(lq_rmil_multi)))\n",
" \n",
" return all_icp_q, all_mil_q, all_rmil_q\n",
"\n",
"\n",
"def mean_errors_all_links_together(Y_w_test):\n",
" \"\"\"Computes mean errors after summing\n",
" predictions from all links.\n",
" \"\"\"\n",
" \n",
" # extract predictions for all links\n",
" Y_true = Y_test[:,0,:,0,0] * Y_scale_test[:,0,:] + Y_rm_mean_test[:,0,:]\n",
" Y_naive = Y_rm_mean_test[:,0,:]\n",
" Y_pred_lr = LR_predictions_mean[0][:,:] * Y_scale_test[:,0,:] + Y_rm_mean_test[:,0,:]\n",
" Y_pred_dl = DL_predictions_mean[0][:,0,:,0,0] * Y_scale_test[:,0,:] + Y_rm_mean_test[:,0,:]\n",
" Y_pred_multi = DL_predictions_multi_q[0][:,0,:,0,0] * Y_scale_test[:,0,:] + Y_rm_mean_test[:,0,:]\n",
"\n",
" # sum predictions over 'links' axis \n",
" Y_true_total = np.sum(Y_true * Y_w_test[:,0,:],axis=1).squeeze()\n",
" Y_naive_total = np.sum(Y_naive * Y_w_test[:,0,:],axis=1).squeeze()\n",
" Y_pred_lr_total = np.sum(Y_pred_lr * Y_w_test[:,0,:],axis=1).squeeze()\n",
" Y_pred_dl_total = np.sum(Y_pred_dl * Y_w_test[:,0,:],axis=1).squeeze()\n",
" Y_pred_multi_total = np.sum(Y_pred_multi * Y_w_test[:,0,:],axis=1).squeeze()\n",
" \n",
" # compute mean errors\n",
" corr_ha, mae_ha, mse_ha, rae_ha, rmse_ha, r2_ha = compute_mean_errors(Y_true_total, Y_naive_total)\n",
" corr_lr, mae_lr, mse_lr, rae_lr, rmse_lr, r2_lr = compute_mean_errors(Y_true_total, Y_pred_lr_total)\n",
" corr_dl, mae_dl, mse_dl, rae_dl, rmse_dl, r2_dl = compute_mean_errors(Y_true_total, Y_pred_dl_total)\n",
" corr_multi, mae_multi, mse_multi, rae_multi, rmse_multi, r2_multi = compute_mean_errors(Y_true_total, Y_pred_multi_total)\n",
"\n",
" print \"MEAN ERRORS (ALL LINKS TOGETHER):\"\n",
" print \"\\tMAE\\tRMSE\\tR2\"\n",
" print \"HA: %.3f %.3f %.3f\" % (mae_ha, rmse_ha, r2_ha) \n",
" print \"LR: %.3f %.3f %.3f\" % (mae_lr, rmse_lr, r2_lr) \n",
" print \"DL: %.3f %.3f %.3f\" % (mae_dl, rmse_dl, r2_dl) \n",
" print \"MULTI: %.3f %.3f %.3f\" % (mae_multi, rmse_multi, r2_multi) \n",
" print \"\\n\"\n",
" \n",
" \n",
"def get_crossing_cases():\n",
" \n",
" num_cross_qr = 0\n",
" num_cross_dl = 0\n",
" num_cross_multi = 0\n",
" num_total = 0\n",
" \n",
" for i in range(num_links):\n",
" for k in range(len(quantiles)-1):\n",
" \n",
" q1 = QR_predictions[k][:,i] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" q2 = QR_predictions[k+1][:,i] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" loss_qr = np.sum(q1 > q2)\n",
" num_cross_qr += loss_qr\n",
" \n",
" q1 = DL_predictions_q[k][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" q2 = DL_predictions_q[k+1][:,0,i,0,0] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" loss_dl = np.sum(q1 > q2)\n",
" num_cross_dl += loss_dl\n",
" \n",
" q1 = DL_predictions_multi_q[0][:,0,i,0,1+k] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" q2 = DL_predictions_multi_q[0][:,0,i,0,1+k+1] * Y_scale_test[:,0,i] + Y_rm_mean_test[:,0,i]\n",
" loss_multi = np.sum(q1 > q2)\n",
" num_cross_multi += loss_multi \n",
" \n",
" num_total += len(q1)\n",
" \n",
" print \"[%d-%d] Crossing Cases: QR: %d\\tDL: %d\\t MULTI: %d\" % (quantiles[k]*100, quantiles[k+1]*100,\n",
" loss_qr,loss_dl,loss_multi)\n",
" \n",
" print \"\\n\"\n",
" print \"Crossing Cases:\"\n",
" print \"QR:\", num_cross_qr \n",
" print \"DL:\", num_cross_dl \n",
" print \"MULTI:\", num_cross_multi\n",
"\n",
" print \"\\n\"\n",
" print \"Crossing Ratios:\"\n",
" print \"QR:\", \"{0:.3}\".format(num_cross_qr / num_total)\n",
" print \"DL:\", \"{0:.3}\".format(num_cross_dl / num_total)\n",
" print \"MULTI:\", \"{0:.3}\".format(num_cross_multi / num_total)\n",
" \n",
"\n",
"def helper(temp):\n",
"\n",
" l2 = []\n",
" for j in range(3):\n",
" l = []\n",
" for i in range(len(temp)):\n",
" l.append(temp[i][j])\n",
" l2.append(l)\n",
" return l2\n",
" \n",
" \n",
"def show_quantile_eval(icp, mil, rmil):\n",
"\n",
" icp = helper(icp)\n",
" mil = helper(mil)\n",
" rmil = helper(rmil)\n",
" \n",
" mets = [icp,mil,rmil]\n",
" \n",
" percentiles= [math.ceil((quantiles[-(k+1)] - quantiles[k])*100) for k in range(int(len(quantiles)/2))][::-1] \n",
" \n",
" fig = plt.figure(figsize=(16,8)) \n",
" labels = ['ICP', 'MIL', 'RMIL']\n",
" for index, met in enumerate(mets):\n",
" ax = fig.add_subplot(111)\n",
" ax.plot(percentiles, met[0][::-1], linewidth=2, marker='o')\n",
" ax.plot(percentiles, met[1][::-1], linewidth=2, marker='o') \n",
" ax.plot(percentiles, met[2][::-1], linewidth=2, marker='o') \n",
" plt.xticks(np.arange(min(percentiles), max(percentiles), (percentiles[-1] - percentiles[-2] - 1)))\n",
" plt.xlabel('Prediction Intervals')\n",
" plt.ylabel(labels[index])\n",
" plt.title('Evolution of ' + labels[index] + ' over Intervals')\n",
" plt.legend(['QR', 'DLQR', 'MULTI'], loc='upper right')\n",
" plt.plot()\n",
" break\n",
" plt.show()\n",
" fig.savefig(\"quantile_eval.pdf\", bbox_inches='tight')\n",
"\n",
" \n",
"def show_loss():\n",
" \n",
" ## loss over epochs, DL model mean\n",
" fig = plt.figure(figsize=(12, 5), dpi=80, facecolor='w', edgecolor='k')\n",
" ax = fig.add_subplot(1,1,1)\n",
" ax.plot(np.array(hist_DL_mean[0].history['loss']) / 60)\n",
" ax.plot(np.array(hist_DL_mean[0].history['val_loss']) / 60, linestyle = '--')\n",
" plt.ylabel('loss (min)')\n",
" plt.xlabel('epochs')\n",
" plt.legend(['train', 'test'], loc='upper right') \n",
" plt.title('DL model for mean')\n",
" plt.plot()\n",
" plt.show()\n",
" fig.savefig(\"dl_mean.pdf\", bbox_inches='tight')\n",
" \n",
" ## loss over epochs, independent DL models for each quantile\n",
" fig = plt.figure(figsize=(16,10))\n",
" fig.subplots_adjust(hspace=0.4, wspace=0.4)\n",
" for i,hist in enumerate(hist_DL_q):\n",
" ax = fig.add_subplot(3, 2, i+1)\n",
" ax.plot(np.array(hist.history['loss']) / 60, label='%s quantile' % str(quantiles[i]))\n",
" ax.plot(np.array(hist.history['val_loss']) / 60, label='%s quantile' % str(quantiles[i]), linestyle = '--')\n",
" plt.ylabel('loss (min)')\n",
" plt.xlabel('epochs')\n",
" plt.legend(['train', 'test'], loc='upper right')\n",
" plt.title('DL model for quantile ' + str(quantiles[i]))\n",
" plt.plot()\n",
" plt.show()\n",
" fig.savefig(\"independent_dl.pdf\", bbox_inches='tight')\n",
" \n",
" ## loss over epochs, DL model for multiple quantiles \n",
" fig = plt.figure(figsize=(12, 5), dpi=80, facecolor='w', edgecolor='k')\n",
" ax = fig.add_subplot(1,1,1)\n",
" ax.plot(np.array(hist_DL_multi_q[0].history['loss']) / 60)\n",
" ax.plot(np.array(hist_DL_multi_q[0].history['val_loss']) / 60, linestyle = '--')\n",
" plt.ylabel('loss (min)')\n",
" plt.xlabel('epochs')\n",
" plt.legend(['train', 'test'], loc='upper right') \n",
" plt.title('DL model for multiple quantiles')\n",
" plt.plot()\n",
" plt.show()\n",
" fig.savefig(\"multi_dl.pdf\", bbox_inches='tight')\n",
"\n",
" \n",
"if __name__ == '__main__':\n",
"\n",
" data = load_data('data/4A_1_201705_201709.csv')\n",
"\n",
" num_links, Y_test, Y_scale_test, Y_rm_mean_test, Y_w_test = models_infer(data)\n",
" mean_errors_all_links_together(Y_w_test)\n",
" icp, mil, rmil = evaluate_quantiles(num_links, Y_test, Y_scale_test, Y_rm_mean_test)\n",
" get_crossing_cases()\n",
" show_quantile_eval(icp, mil, rmil)\n",
" show_loss()\n",
" "
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [default]",
"language": "python",
"name": "python2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.14"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment