conf_figures.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: 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"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import json\n",
    "import pickle\n",
    "import random\n",
    "import numpy as np\n",
    "import itertools\n",
    "import argparse\n",
    "from tqdm import tqdm, tqdm_notebook\n",
    "from plotnine import *\n",
    "import pandas as pd\n",
    "\n",
    "import cupy\n",
    "import chainer\n",
    "from chainer import functions as F\n",
    "\n",
    "from nlp_utils import convert_seq, convert_snli_seq\n",
    "from utils import setup_model\n",
    "\n",
    "from run_dknn import DkNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"gpu\": 0,\n",
      "  \"model_setup\": \"result/snli_bilstm/args.json\",\n",
      "  \"lsh\": false\n",
      "}\n",
      "{\n",
      "  \"batchsize\": 128,\n",
      "  \"epoch\": 10,\n",
      "  \"gpu\": 0,\n",
      "  \"out\": \"result\",\n",
      "  \"unit\": 300,\n",
      "  \"layer\": 3,\n",
      "  \"dropout\": 0.4,\n",
      "  \"combine_snli\": false,\n",
      "  \"dataset\": \"snli\",\n",
      "  \"model\": \"bilstm\",\n",
      "  \"char_based\": false,\n",
      "  \"word_vectors\": null,\n",
      "  \"save_path\": \"/home/t-shfeng/deep-knn/result/snli_bilstm\",\n",
      "  \"vocab_path\": \"/home/t-shfeng/deep-knn/result/snli_bilstm/vocab.json\",\n",
      "  \"model_path\": \"/home/t-shfeng/deep-knn/result/snli_bilstm/best_model.npz\",\n",
      "  \"n_class\": 3,\n",
      "  \"datetime\": \"2018-07-09 07:48:46.853809\"\n",
      "}\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "download snli\n",
      "read snli\n",
      "construct vocabulary based on frequency\n",
      "# train data: 550152\n",
      "# test  data: 10000\n",
      "# vocab: 39431\n",
      "# class: 3\n"
     ]
    }
   ],
   "source": [
    "parser = argparse.ArgumentParser()\n",
    "parser.add_argument('--gpu', '-g', type=int, default=0,\n",
    "                    help='GPU ID (negative value indicates CPU)')\n",
    "parser.add_argument('--model-setup', required=True,\n",
    "                    help='Model setup dictionary.')\n",
    "parser.add_argument('--lsh', action='store_true', default=False,\n",
    "                    help='If true, uses locally sensitive hashing \\\n",
    "                          (with k=10 NN) for NN search.')\n",
    "args = parser.parse_args(['--model-setup', 'result/snli_bilstm/args.json'])\n",
    "\n",
    "model, train, test, vocab, setup = setup_model(args)\n",
    "if setup['dataset'] == 'snli' and not setup['combine_snli']:\n",
    "    converter = convert_snli_seq\n",
    "else:\n",
    "    converter = convert_seq\n",
    "\n",
    "# FIXME\n",
    "args.batchsize = 64\n",
    "max_beam_size = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(os.path.join(setup['save_path'], 'calib.json')) as f:\n",
    "    calibration_idx = json.load(f)\n",
    "\n",
    "calibration = [train[i] for i in calibration_idx]\n",
    "train = [x for i, x in enumerate(train) if i not in calibration_idx]\n",
    "train = random.sample(train, 10000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|▍         | 3/78 [00:00<00:03, 20.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "caching hiddens\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "79it [00:03, 20.61it/s]                        \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using KDTree for NN Search\n",
      "building tree for layer 0\n",
      "building tree for layer 1\n",
      "building tree for layer 2\n",
      "building tree for layer 3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/7 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "calibrating credibility\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "8it [01:07,  8.44s/it]                       \n"
     ]
    }
   ],
   "source": [
    "dknn = DkNN(model)\n",
    "dknn.build(train, batch_size=setup['batchsize'], converter=converter, device=args.gpu)\n",
    "dknn.calibrate(calibration, batch_size=setup['batchsize'], converter=converter, device=args.gpu)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "conf = {'confidence': [], 'data': [], 'label': [], 'type': []}\n",
    "dis = {'distance': [], 'data': [], 'label': []}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "ckp = pickle.load(open('rawr_dev.pkl', 'rb'))\n",
    "rd_test =  [x[0]['reduced_input'] for x in ckp]\n",
    "\n",
    "test = test\n",
    "rd_test = rd_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4aa3ef6435b44bba9421c53d6415b478",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=156), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "og_iter = chainer.iterators.SerialIterator(test, args.batchsize, repeat=False, shuffle=False)\n",
    "\n",
    "n_batches = len(test) // args.batchsize\n",
    "for batch in tqdm_notebook(og_iter, total=n_batches):\n",
    "    batch = converter(batch, device=args.gpu)\n",
    "    knn_pred, knn_cred, knn_conf, reg_pred, reg_conf = dknn.predict(batch['xs'], calibrated=True, snli=True)\n",
    "    conf['confidence'].extend(reg_conf)\n",
    "    conf['data'].extend('original' for _ in reg_conf)\n",
    "    conf['label'].extend(reg_pred)\n",
    "    conf['type'].extend('reg_conf' for _ in reg_conf)\n",
    "    \n",
    "    conf['confidence'].extend(knn_conf)\n",
    "    conf['data'].extend('original' for _ in knn_conf)\n",
    "    conf['label'].extend(knn_pred)\n",
    "    conf['type'].extend('knn_conf' for _ in knn_conf)\n",
    "    \n",
    "    conf['confidence'].extend(knn_cred)\n",
    "    conf['data'].extend('original' for _ in knn_cred)\n",
    "    conf['label'].extend(knn_pred)\n",
    "    conf['type'].extend('knn_cred' for _ in knn_cred)\n",
    "    \n",
    "    ds = dknn.get_nearest_distance(batch['xs'], layer_id=0)\n",
    "    dis['distance'].extend(ds)\n",
    "    dis['data'].extend('original' for _ in ds)\n",
    "    dis['label'].extend(knn_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9980a672fa524ef1b9ba2650f4d47bf1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=156), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "rd_iter = chainer.iterators.SerialIterator(rd_test, args.batchsize, repeat=False, shuffle=False)\n",
    "\n",
    "n_batches = len(rd_test) // args.batchsize\n",
    "for batch in tqdm_notebook(rd_iter, total=n_batches):\n",
    "    batch = converter(batch, device=args.gpu, with_label=False)\n",
    "    knn_pred, knn_cred, knn_conf, reg_pred, reg_conf = dknn.predict(batch, calibrated=True, snli=True)\n",
    "    conf['confidence'].extend(reg_conf)\n",
    "    conf['data'].extend('reduced' for _ in reg_conf)\n",
    "    conf['label'].extend(reg_pred)\n",
    "    conf['type'].extend('reg_conf' for _ in reg_conf)\n",
    "    \n",
    "    conf['confidence'].extend(knn_conf)\n",
    "    conf['data'].extend('reduced' for _ in knn_conf)\n",
    "    conf['label'].extend(knn_pred)\n",
    "    conf['type'].extend('knn_conf' for _ in knn_conf)\n",
    "    \n",
    "    conf['confidence'].extend(knn_cred)\n",
    "    conf['data'].extend('reduced' for _ in knn_cred)\n",
    "    conf['label'].extend(knn_pred)\n",
    "    conf['type'].extend('knn_cred' for _ in knn_cred)\n",
    "    \n",
    "    ds = dknn.get_nearest_distance(batch, layer_id=0)\n",
    "    dis['distance'].extend(ds)\n",
    "    dis['data'].extend('reduced' for _ in ds)\n",
    "    dis['label'].extend(knn_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "conf = pd.DataFrame(conf)\n",
    "dis = pd.DataFrame(dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/plotnine/utils.py:281: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  ndistinct = ids.apply(len_unique, axis=0).as_matrix()\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:4384: FutureWarning: Attribute 'is_copy' is deprecated and will be removed in a future version.\n",
      "  object.__getattribute__(self, name)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:4385: FutureWarning: Attribute 'is_copy' is deprecated and will be removed in a future version.\n",
      "  return object.__setattr__(self, name, value)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/statsmodels/sandbox/nonparametric/kernels.py:204: RuntimeWarning: divide by zero encountered in double_scalars\n",
      "  w = 1. / (h * n) * np.sum(self((xs-x)/h), axis=0)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/statsmodels/sandbox/nonparametric/kernels.py:204: RuntimeWarning: divide by zero encountered in true_divide\n",
      "  w = 1. / (h * n) * np.sum(self((xs-x)/h), axis=0)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/statsmodels/sandbox/nonparametric/kernels.py:204: RuntimeWarning: invalid value encountered in true_divide\n",
      "  w = 1. / (h * n) * np.sum(self((xs-x)/h), axis=0)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/statsmodels/sandbox/nonparametric/kernels.py:204: RuntimeWarning: invalid value encountered in multiply\n",
      "  w = 1. / (h * n) * np.sum(self((xs-x)/h), axis=0)\n"
     ]
    },
    {
     "ename": "ValueError",
     "evalue": "zero-size array to reduction operation maximum which has no identity",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/IPython/core/formatters.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    700\u001b[0m                 \u001b[0mtype_pprinters\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtype_printers\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    701\u001b[0m                 deferred_pprinters=self.deferred_printers)\n\u001b[0;32m--> 702\u001b[0;31m             \u001b[0mprinter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpretty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    703\u001b[0m             \u001b[0mprinter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mflush\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    704\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mstream\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetvalue\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/IPython/lib/pretty.py\u001b[0m in \u001b[0;36mpretty\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    398\u001b[0m                         \u001b[0;32mif\u001b[0m \u001b[0mcls\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mobject\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    399\u001b[0m                                 \u001b[0;32mand\u001b[0m \u001b[0mcallable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__dict__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'__repr__'\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;32m--> 400\u001b[0;31m                             \u001b[0;32mreturn\u001b[0m \u001b[0m_repr_pprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcycle\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    401\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    402\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0m_default_pprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcycle\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/IPython/lib/pretty.py\u001b[0m in \u001b[0;36m_repr_pprint\u001b[0;34m(obj, p, cycle)\u001b[0m\n\u001b[1;32m    693\u001b[0m     \u001b[0;34m\"\"\"A pprint that just redirects to the normal repr function.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    694\u001b[0m     \u001b[0;31m# Find newlines and replace them with p.break_()\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 695\u001b[0;31m     \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrepr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    696\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0moutput_line\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplitlines\u001b[0m\u001b[0;34m(\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[1;32m    697\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/ggplot.py\u001b[0m in \u001b[0;36m__repr__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     84\u001b[0m         \u001b[0mPrint\u001b[0m\u001b[0;34m/\u001b[0m\u001b[0mshow\u001b[0m \u001b[0mthe\u001b[0m \u001b[0mplot\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     85\u001b[0m         \"\"\"\n\u001b[0;32m---> 86\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw\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     87\u001b[0m         \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     88\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0;34m'<ggplot: (%d)>'\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__hash__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/ggplot.py\u001b[0m in \u001b[0;36mdraw\u001b[0;34m(self, return_ggplot)\u001b[0m\n\u001b[1;32m    177\u001b[0m         \u001b[0;31m# assign a default theme\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    178\u001b[0m         \u001b[0mself\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdeepcopy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 179\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_build\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    180\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    181\u001b[0m         \u001b[0;31m# If no theme we use the default\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/ggplot.py\u001b[0m in \u001b[0;36m_build\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    285\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    286\u001b[0m         \u001b[0;31m# Apply and map statistics\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 287\u001b[0;31m         \u001b[0mlayers\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_statistic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlayout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    288\u001b[0m         \u001b[0mlayers\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmap_statistic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    289\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/layer.py\u001b[0m in \u001b[0;36mcompute_statistic\u001b[0;34m(self, layout)\u001b[0m\n\u001b[1;32m     84\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mcompute_statistic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlayout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     85\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0ml\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 86\u001b[0;31m             \u001b[0ml\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_statistic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlayout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     87\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     88\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mmap_statistic\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mplot\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/layer.py\u001b[0m in \u001b[0;36mcompute_statistic\u001b[0;34m(self, layout)\u001b[0m\n\u001b[1;32m    361\u001b[0m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0muse_defaults\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    362\u001b[0m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msetup_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 363\u001b[0;31m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_layer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mparams\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlayout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    364\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    365\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/stats/stat.py\u001b[0m in \u001b[0;36mcompute_layer\u001b[0;34m(cls, data, params, layout)\u001b[0m\n\u001b[1;32m    201\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_panel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpscales\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    202\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 203\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mgroupby_apply\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'PANEL'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    204\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    205\u001b[0m     \u001b[0;34m@\u001b[0m\u001b[0mclassmethod\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/utils.py\u001b[0m in \u001b[0;36mgroupby_apply\u001b[0;34m(df, cols, func, *args, **kwargs)\u001b[0m\n\u001b[1;32m    612\u001b[0m         \u001b[0;31m# do not mark d as a slice of df i.e no SettingWithCopyWarning\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    613\u001b[0m         \u001b[0md\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_copy\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 614\u001b[0;31m         \u001b[0mlst\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0md\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    615\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconcat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlst\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mignore_index\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    616\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/stats/stat.py\u001b[0m in \u001b[0;36mfn\u001b[0;34m(pdata)\u001b[0m\n\u001b[1;32m    199\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mpdata\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    200\u001b[0m             \u001b[0mpscales\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlayout\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_scales\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'PANEL'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miat\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 201\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_panel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpscales\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    202\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mgroupby_apply\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'PANEL'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/stats/stat.py\u001b[0m in \u001b[0;36mcompute_panel\u001b[0;34m(cls, data, scales, **params)\u001b[0m\n\u001b[1;32m    228\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0m_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mold\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgroupby\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'group'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    229\u001b[0m             \u001b[0mold\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_copy\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 230\u001b[0;31m             \u001b[0mnew\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompute_group\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mold\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscales\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    231\u001b[0m             \u001b[0munique\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0muniquecols\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mold\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    232\u001b[0m             \u001b[0mmissing\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0munique\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdifference\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnew\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/stats/stat_density.py\u001b[0m in \u001b[0;36mcompute_group\u001b[0;34m(cls, data, scales, **params)\u001b[0m\n\u001b[1;32m    130\u001b[0m             \u001b[0mrange_x\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mscales\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdimension\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    131\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 132\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mcompute_density\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'x'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrange_x\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    133\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    134\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/plotnine/stats/stat_density.py\u001b[0m in \u001b[0;36mcompute_density\u001b[0;34m(x, weight, range, **params)\u001b[0m\n\u001b[1;32m    197\u001b[0m     return pd.DataFrame({'x': x2,\n\u001b[1;32m    198\u001b[0m                          \u001b[0;34m'density'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m                          \u001b[0;34m'scaled'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my\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    200\u001b[0m                          \u001b[0;34m'count'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    201\u001b[0m                          'n': n})\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py\u001b[0m in \u001b[0;36mamax\u001b[0;34m(a, axis, out, keepdims)\u001b[0m\n\u001b[1;32m   2318\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2319\u001b[0m     return _methods._amax(a, axis=axis,\n\u001b[0;32m-> 2320\u001b[0;31m                           out=out, **kwargs)\n\u001b[0m\u001b[1;32m   2321\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2322\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/numpy/core/_methods.py\u001b[0m in \u001b[0;36m_amax\u001b[0;34m(a, axis, out, keepdims)\u001b[0m\n\u001b[1;32m     24\u001b[0m \u001b[0;31m# small reductions\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     25\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_amax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkeepdims\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mumr_maximum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkeepdims\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     27\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     28\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_amin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkeepdims\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: zero-size array to reduction operation maximum which has no identity"
     ]
    }
   ],
   "source": [
    "(\n",
    "    ggplot(conf) +\n",
    "    aes(x='confidence', color='data', fill='data') +\n",
    "    geom_density(alpha=.45) +\n",
    "    facet_grid('type ~ label')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/plotnine/utils.py:281: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  ndistinct = ids.apply(len_unique, axis=0).as_matrix()\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:4384: FutureWarning: Attribute 'is_copy' is deprecated and will be removed in a future version.\n",
      "  object.__getattribute__(self, name)\n",
      "/home/t-shfeng/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py:4385: FutureWarning: Attribute 'is_copy' is deprecated and will be removed in a future version.\n",
      "  return object.__setattr__(self, name, value)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (-9223363261982951196)>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    ggplot(dis) +\n",
    "    aes(x='distance', color='data', fill='data') +\n",
    "    geom_density(alpha=0.45) + \n",
    "    facet_wrap('label')\n",
    ")"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}