|
m@65
|
1 {
|
|
m@65
|
2 "cells": [
|
|
m@65
|
3 {
|
|
m@65
|
4 "cell_type": "code",
|
|
m@71
|
5 "execution_count": 1,
|
|
m@79
|
6 "metadata": {},
|
|
m@65
|
7 "outputs": [
|
|
m@65
|
8 {
|
|
m@71
|
9 "name": "stderr",
|
|
m@65
|
10 "output_type": "stream",
|
|
m@65
|
11 "text": [
|
|
m@71
|
12 "/homes/mp305/anaconda/lib/python2.7/site-packages/librosa/core/audio.py:33: UserWarning: Could not import scikits.samplerate. Falling back to scipy.signal\n",
|
|
m@71
|
13 " warnings.warn('Could not import scikits.samplerate. '\n"
|
|
m@65
|
14 ]
|
|
m@65
|
15 }
|
|
m@65
|
16 ],
|
|
m@65
|
17 "source": [
|
|
m@65
|
18 "import numpy as np\n",
|
|
m@65
|
19 "import pandas as pd\n",
|
|
m@65
|
20 "import pickle \n",
|
|
m@65
|
21 "\n",
|
|
m@65
|
22 "%load_ext autoreload\n",
|
|
m@65
|
23 "%autoreload 2\n",
|
|
m@65
|
24 "\n",
|
|
m@65
|
25 "%matplotlib inline\n",
|
|
m@65
|
26 "import matplotlib.pyplot as plt\n",
|
|
m@65
|
27 "\n",
|
|
m@65
|
28 "import sys\n",
|
|
m@65
|
29 "sys.path.append('../')\n",
|
|
m@65
|
30 "import scripts.load_dataset as load_dataset\n",
|
|
m@65
|
31 "import scripts.map_and_average as mapper\n",
|
|
m@65
|
32 "import scripts.classification as classification\n",
|
|
m@65
|
33 "import scripts.outliers as outliers\n",
|
|
m@65
|
34 "import scripts.utils as utils"
|
|
m@65
|
35 ]
|
|
m@65
|
36 },
|
|
m@65
|
37 {
|
|
m@65
|
38 "cell_type": "code",
|
|
m@65
|
39 "execution_count": 2,
|
|
m@65
|
40 "metadata": {
|
|
m@65
|
41 "collapsed": true
|
|
m@65
|
42 },
|
|
m@65
|
43 "outputs": [],
|
|
m@65
|
44 "source": [
|
|
m@65
|
45 "def limit_to_n_seconds(dataset, n_sec=30.0, win_sec=8.0):\n",
|
|
m@65
|
46 " X, Y, Yaudio = dataset\n",
|
|
m@65
|
47 " uniq_audio, uniq_counts = np.unique(Yaudio, return_counts=True)\n",
|
|
m@65
|
48 " frame_sr = 2.0\n",
|
|
m@65
|
49 " max_n_frames = np.int(np.floor((n_sec - win_sec) * frame_sr))\n",
|
|
m@65
|
50 " X_new, Y_new, Yaudio_new = [], [], []\n",
|
|
m@65
|
51 " for audio in uniq_audio:\n",
|
|
m@65
|
52 " idx = np.where(Yaudio==audio)[0]\n",
|
|
m@65
|
53 " if len(idx) > max_n_frames:\n",
|
|
m@65
|
54 " idx = idx[:max_n_frames]\n",
|
|
m@65
|
55 " X_new.append(X[idx, :])\n",
|
|
m@65
|
56 " Y_new.append(Y[idx])\n",
|
|
m@65
|
57 " Yaudio_new.append(Yaudio[idx])\n",
|
|
m@65
|
58 " return [np.concatenate(X_new), np.concatenate(Y_new), np.concatenate(Yaudio_new)]"
|
|
m@65
|
59 ]
|
|
m@65
|
60 },
|
|
m@65
|
61 {
|
|
m@65
|
62 "cell_type": "code",
|
|
m@65
|
63 "execution_count": null,
|
|
m@65
|
64 "metadata": {
|
|
m@65
|
65 "collapsed": true
|
|
m@65
|
66 },
|
|
m@65
|
67 "outputs": [],
|
|
m@65
|
68 "source": [
|
|
m@65
|
69 "trainset, valset, testset = mapper.load_train_val_test_sets()\n",
|
|
m@65
|
70 "trainset = limit_to_n_seconds(trainset)\n",
|
|
m@65
|
71 "valset = limit_to_n_seconds(valset)\n",
|
|
m@65
|
72 "testset = limit_to_n_seconds(testset)"
|
|
m@65
|
73 ]
|
|
m@65
|
74 },
|
|
m@65
|
75 {
|
|
m@65
|
76 "cell_type": "code",
|
|
m@65
|
77 "execution_count": null,
|
|
m@65
|
78 "metadata": {
|
|
m@65
|
79 "collapsed": true
|
|
m@65
|
80 },
|
|
m@65
|
81 "outputs": [],
|
|
m@65
|
82 "source": [
|
|
m@65
|
83 "print np.array_equal(np.unique(trainset[1]), np.unique(valset[1]))\n",
|
|
m@65
|
84 "print np.array_equal(np.unique(testset[1]), np.unique(valset[1]))\n",
|
|
m@65
|
85 "print np.array_equal(np.unique(testset[1]), np.unique(trainset[1]))"
|
|
m@65
|
86 ]
|
|
m@65
|
87 },
|
|
m@65
|
88 {
|
|
m@65
|
89 "cell_type": "code",
|
|
m@79
|
90 "execution_count": 2,
|
|
m@79
|
91 "metadata": {},
|
|
m@65
|
92 "outputs": [
|
|
m@65
|
93 {
|
|
m@65
|
94 "name": "stdout",
|
|
m@65
|
95 "output_type": "stream",
|
|
m@65
|
96 "text": [
|
|
m@65
|
97 "['/import/c4dm-04/mariap/train_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/val_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/test_data_melodia_8_30sec.pickle']\n"
|
|
m@65
|
98 ]
|
|
m@65
|
99 }
|
|
m@65
|
100 ],
|
|
m@65
|
101 "source": [
|
|
m@65
|
102 "OUTPUT_FILES = [output_file.split('.pickle')[0]+'_30sec.pickle' for \n",
|
|
m@65
|
103 " output_file in load_dataset.OUTPUT_FILES]\n",
|
|
m@65
|
104 "print OUTPUT_FILES"
|
|
m@65
|
105 ]
|
|
m@65
|
106 },
|
|
m@65
|
107 {
|
|
m@65
|
108 "cell_type": "code",
|
|
m@65
|
109 "execution_count": 12,
|
|
m@65
|
110 "metadata": {
|
|
m@65
|
111 "collapsed": true
|
|
m@65
|
112 },
|
|
m@65
|
113 "outputs": [],
|
|
m@65
|
114 "source": [
|
|
m@65
|
115 "if 1:\n",
|
|
m@65
|
116 " with open(OUTPUT_FILES[0], 'wb') as f:\n",
|
|
m@65
|
117 " pickle.dump(trainset, f) \n",
|
|
m@65
|
118 " with open(OUTPUT_FILES[1], 'wb') as f:\n",
|
|
m@65
|
119 " pickle.dump(valset, f)\n",
|
|
m@65
|
120 " with open(OUTPUT_FILES[2], 'wb') as f:\n",
|
|
m@65
|
121 " pickle.dump(testset, f)"
|
|
m@65
|
122 ]
|
|
m@65
|
123 },
|
|
m@65
|
124 {
|
|
m@65
|
125 "cell_type": "code",
|
|
m@79
|
126 "execution_count": 3,
|
|
m@79
|
127 "metadata": {},
|
|
m@65
|
128 "outputs": [
|
|
m@65
|
129 {
|
|
m@65
|
130 "name": "stdout",
|
|
m@65
|
131 "output_type": "stream",
|
|
m@65
|
132 "text": [
|
|
m@71
|
133 "['/import/c4dm-04/mariap/train_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/val_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/test_data_melodia_8_30sec.pickle'] ['/import/c4dm-04/mariap/lda_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/pca_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/nmf_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/ssnmf_data_melodia_8_30sec.pickle', '/import/c4dm-04/mariap/na_data_melodia_8_30sec.pickle']\n"
|
|
m@65
|
134 ]
|
|
m@65
|
135 }
|
|
m@65
|
136 ],
|
|
m@65
|
137 "source": [
|
|
m@65
|
138 "mapper.INPUT_FILES = OUTPUT_FILES\n",
|
|
m@65
|
139 "MAPPER_OUTPUT_FILES = mapper.OUTPUT_FILES\n",
|
|
m@65
|
140 "mapper.OUTPUT_FILES = [output_file.split('.pickle')[0]+'_30sec.pickle' for \n",
|
|
m@65
|
141 " output_file in MAPPER_OUTPUT_FILES]\n",
|
|
m@65
|
142 "print mapper.INPUT_FILES, mapper.OUTPUT_FILES"
|
|
m@65
|
143 ]
|
|
m@65
|
144 },
|
|
m@65
|
145 {
|
|
m@65
|
146 "cell_type": "code",
|
|
m@71
|
147 "execution_count": null,
|
|
m@79
|
148 "metadata": {},
|
|
m@65
|
149 "outputs": [
|
|
m@65
|
150 {
|
|
m@65
|
151 "name": "stdout",
|
|
m@65
|
152 "output_type": "stream",
|
|
m@65
|
153 "text": [
|
|
m@65
|
154 "mapping...\n",
|
|
m@65
|
155 "(209292, 840) (70087, 840) (69766, 840)\n",
|
|
m@65
|
156 "mapping rhy\n",
|
|
m@65
|
157 "training with PCA transform...\n",
|
|
m@65
|
158 "variance explained 1.0\n",
|
|
m@65
|
159 "138 400\n",
|
|
m@65
|
160 "training with PCA transform...\n",
|
|
m@71
|
161 "variance explained 0.989994197011\n",
|
|
m@65
|
162 "training with LDA transform...\n"
|
|
m@65
|
163 ]
|
|
m@65
|
164 },
|
|
m@65
|
165 {
|
|
m@65
|
166 "name": "stderr",
|
|
m@65
|
167 "output_type": "stream",
|
|
m@65
|
168 "text": [
|
|
m@71
|
169 "/homes/mp305/anaconda/lib/python2.7/site-packages/sklearn/utils/validation.py:526: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
|
|
m@71
|
170 " y = column_or_1d(y, warn=True)\n",
|
|
m@65
|
171 "/homes/mp305/anaconda/lib/python2.7/site-packages/sklearn/discriminant_analysis.py:455: UserWarning: The priors do not sum to 1. Renormalizing\n",
|
|
m@65
|
172 " UserWarning)\n"
|
|
m@65
|
173 ]
|
|
m@65
|
174 },
|
|
m@65
|
175 {
|
|
m@65
|
176 "name": "stdout",
|
|
m@65
|
177 "output_type": "stream",
|
|
m@65
|
178 "text": [
|
|
m@65
|
179 "variance explained 1.0\n",
|
|
m@71
|
180 "training with NMF transform...\n",
|
|
m@71
|
181 "reconstruction error 6.59195506061\n",
|
|
m@71
|
182 "training with SSNMF transform...\n",
|
|
m@71
|
183 "reconstruction error 25.0727210368\n",
|
|
m@65
|
184 "transform test data...\n",
|
|
m@65
|
185 "mapping mel\n",
|
|
m@65
|
186 "training with PCA transform...\n",
|
|
m@65
|
187 "variance explained 1.0\n",
|
|
m@65
|
188 "214 240\n",
|
|
m@65
|
189 "training with PCA transform...\n",
|
|
m@65
|
190 "variance explained 0.990347897477\n",
|
|
m@65
|
191 "training with LDA transform...\n",
|
|
m@65
|
192 "variance explained 1.0\n",
|
|
m@71
|
193 "training with NMF transform...\n"
|
|
m@65
|
194 ]
|
|
m@65
|
195 }
|
|
m@65
|
196 ],
|
|
m@65
|
197 "source": [
|
|
m@65
|
198 "print \"mapping...\"\n",
|
|
m@71
|
199 "#_, _, ldadata_list, _, _, Y, Yaudio = mapper.lda_map_and_average_frames(min_variance=0.99)\n",
|
|
m@71
|
200 "#mapper.write_output([], [], ldadata_list, [], [], Y, Yaudio)\n",
|
|
m@71
|
201 "data_list, pcadata_list, ldadata_list, nmfdata_list, ssnmfdata_list, classlabs, audiolabs = mapper.map_and_average_frames(min_variance=0.99)\n",
|
|
m@71
|
202 "mapper.write_output(data_list, pcadata_list, ldadata_list, nmfdata_list, ssnmfdata_list, classlabs, audiolabs)"
|
|
m@65
|
203 ]
|
|
m@65
|
204 },
|
|
m@65
|
205 {
|
|
m@65
|
206 "cell_type": "code",
|
|
m@82
|
207 "execution_count": 4,
|
|
m@65
|
208 "metadata": {
|
|
m@65
|
209 "scrolled": true
|
|
m@65
|
210 },
|
|
m@65
|
211 "outputs": [
|
|
m@65
|
212 {
|
|
m@65
|
213 "name": "stdout",
|
|
m@65
|
214 "output_type": "stream",
|
|
m@65
|
215 "text": [
|
|
m@82
|
216 "/import/c4dm-04/mariap/lda_data_melodia_8_30sec.pickle\n"
|
|
m@82
|
217 ]
|
|
m@82
|
218 },
|
|
m@82
|
219 {
|
|
m@82
|
220 "name": "stderr",
|
|
m@82
|
221 "output_type": "stream",
|
|
m@82
|
222 "text": [
|
|
m@82
|
223 "/homes/mp305/anaconda/lib/python2.7/site-packages/sklearn/metrics/classification.py:1113: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.\n",
|
|
m@82
|
224 " 'precision', 'predicted', average, warn_for)\n",
|
|
m@82
|
225 "/homes/mp305/anaconda/lib/python2.7/site-packages/sklearn/discriminant_analysis.py:455: UserWarning: The priors do not sum to 1. Renormalizing\n",
|
|
m@82
|
226 " UserWarning)\n"
|
|
m@82
|
227 ]
|
|
m@82
|
228 },
|
|
m@82
|
229 {
|
|
m@82
|
230 "name": "stdout",
|
|
m@82
|
231 "output_type": "stream",
|
|
m@82
|
232 "text": [
|
|
m@79
|
233 "KNN LDA 0.151978449974\n",
|
|
m@79
|
234 "LDA LDA 0.320669835863\n",
|
|
m@79
|
235 "SVM LDA 0.0231101788399\n",
|
|
m@82
|
236 "RF LDA 0.0750711341437\n",
|
|
m@79
|
237 "KNN LDA 0.0547390436205\n",
|
|
m@79
|
238 "LDA LDA 0.150312531138\n",
|
|
m@79
|
239 "SVM LDA 0.0787628988868\n",
|
|
m@82
|
240 "RF LDA 0.0325169847578\n",
|
|
m@79
|
241 "KNN LDA 0.0232330458268\n",
|
|
m@79
|
242 "LDA LDA 0.0702474072041\n",
|
|
m@79
|
243 "SVM LDA 0.050068706152\n",
|
|
m@82
|
244 "RF LDA 0.0221962788765\n",
|
|
m@79
|
245 "KNN LDA 0.281733731607\n",
|
|
m@79
|
246 "LDA LDA 0.198582742899\n",
|
|
m@79
|
247 "SVM LDA 0.296355560166\n",
|
|
m@82
|
248 "RF LDA 0.138929482109\n",
|
|
m@79
|
249 "KNN LDA 0.0857923493684\n",
|
|
m@79
|
250 "LDA LDA 0.107355289483\n",
|
|
m@79
|
251 "SVM LDA 0.0896098014444\n",
|
|
m@82
|
252 "RF LDA 0.0520582564401\n",
|
|
m@79
|
253 "/import/c4dm-04/mariap/pca_data_melodia_8_30sec.pickle\n",
|
|
m@79
|
254 "KNN PCA 0.140643930221\n",
|
|
m@79
|
255 "LDA PCA 0.175099072208\n",
|
|
m@79
|
256 "SVM PCA 0.0149273059799\n",
|
|
m@82
|
257 "RF PCA 0.0436883799804\n",
|
|
m@79
|
258 "KNN PCA 0.052516908106\n",
|
|
m@79
|
259 "LDA PCA 0.055028942176\n",
|
|
m@79
|
260 "SVM PCA 0.0479512645907\n",
|
|
m@82
|
261 "RF PCA 0.0293191117144\n",
|
|
m@79
|
262 "KNN PCA 0.0268729640269\n",
|
|
m@79
|
263 "LDA PCA 0.0459303318699\n",
|
|
m@79
|
264 "SVM PCA 0.0386730267598\n",
|
|
m@82
|
265 "RF PCA 0.0200505661024\n",
|
|
m@79
|
266 "KNN PCA 0.220850433533\n",
|
|
m@79
|
267 "LDA PCA 0.161502657527\n",
|
|
m@79
|
268 "SVM PCA 0.245790916558\n",
|
|
m@82
|
269 "RF PCA 0.109652967294\n",
|
|
m@79
|
270 "KNN PCA 0.0814272808267\n",
|
|
m@79
|
271 "LDA PCA 0.0839732813486\n",
|
|
m@79
|
272 "SVM PCA 0.0918638232782\n",
|
|
m@82
|
273 "RF PCA 0.0512694004704\n",
|
|
m@79
|
274 "/import/c4dm-04/mariap/nmf_data_melodia_8_30sec.pickle\n",
|
|
m@79
|
275 "KNN NMF 0.114298949339\n",
|
|
m@79
|
276 "LDA NMF 0.178244078869\n",
|
|
m@79
|
277 "SVM NMF 0.0164055663008\n",
|
|
m@82
|
278 "RF NMF 0.0603986437872\n",
|
|
m@79
|
279 "KNN NMF 0.043057794756\n",
|
|
m@79
|
280 "LDA NMF 0.0586662842996\n",
|
|
m@79
|
281 "SVM NMF 0.00781273342686\n",
|
|
m@82
|
282 "RF NMF 0.0394888564357\n",
|
|
m@79
|
283 "KNN NMF 0.0285281454673\n",
|
|
m@79
|
284 "LDA NMF 0.0463659955869\n",
|
|
m@79
|
285 "SVM NMF 0.00768887594564\n",
|
|
m@82
|
286 "RF NMF 0.0243236536555\n",
|
|
m@79
|
287 "KNN NMF 0.177819886656\n",
|
|
m@79
|
288 "LDA NMF 0.166221515627\n",
|
|
m@79
|
289 "SVM NMF 0.010788613595\n",
|
|
m@82
|
290 "RF NMF 0.125063096463\n",
|
|
m@79
|
291 "KNN NMF 0.0795454671166\n",
|
|
m@79
|
292 "LDA NMF 0.0856428557896\n",
|
|
m@79
|
293 "SVM NMF 0.0116920633048\n",
|
|
m@82
|
294 "RF NMF 0.0410278341205\n",
|
|
m@79
|
295 "/import/c4dm-04/mariap/ssnmf_data_melodia_8_30sec.pickle\n",
|
|
m@79
|
296 "KNN SSNMF 0.14322692821\n",
|
|
m@79
|
297 "LDA SSNMF 0.18320247367\n",
|
|
m@79
|
298 "SVM SSNMF 0.0205784326384\n",
|
|
m@82
|
299 "RF SSNMF 0.0459476869017\n",
|
|
m@79
|
300 "KNN SSNMF 0.0431300683181\n",
|
|
m@79
|
301 "LDA SSNMF 0.0533449581285\n",
|
|
m@79
|
302 "SVM SSNMF 0.0106542141335\n",
|
|
m@82
|
303 "RF SSNMF 0.0267217709348\n",
|
|
m@79
|
304 "KNN SSNMF 0.0152235481009\n",
|
|
m@79
|
305 "LDA SSNMF 0.038872838043\n",
|
|
m@79
|
306 "SVM SSNMF 0.00536127803533\n",
|
|
m@82
|
307 "RF SSNMF 0.0196169872436\n",
|
|
m@79
|
308 "KNN SSNMF 0.227101074174\n",
|
|
m@79
|
309 "LDA SSNMF 0.165382484171\n",
|
|
m@79
|
310 "SVM SSNMF 0.0184921176111\n",
|
|
m@82
|
311 "RF SSNMF 0.0924443700566\n",
|
|
m@79
|
312 "KNN SSNMF 0.0715413500709\n",
|
|
m@79
|
313 "LDA SSNMF 0.0819764377219\n",
|
|
m@79
|
314 "SVM SSNMF 0.0138822224913\n",
|
|
m@82
|
315 "RF SSNMF 0.0347436738199\n",
|
|
m@79
|
316 "/import/c4dm-04/mariap/na_data_melodia_8_30sec.pickle\n",
|
|
m@79
|
317 "KNN NA 0.140075287804\n",
|
|
m@79
|
318 "LDA NA 0.176953549195\n",
|
|
m@79
|
319 "SVM NA 0.0149485545637\n",
|
|
m@82
|
320 "RF NA 0.0681836566939\n",
|
|
m@79
|
321 "KNN NA 0.0515315452955\n",
|
|
m@79
|
322 "LDA NA 0.0599453579616\n",
|
|
m@79
|
323 "SVM NA 0.0468615478392\n",
|
|
m@82
|
324 "RF NA 0.045878758841\n",
|
|
m@79
|
325 "KNN NA 0.0273364752119\n",
|
|
m@79
|
326 "LDA NA 0.0378819151174\n",
|
|
m@79
|
327 "SVM NA 0.038290667129\n",
|
|
m@82
|
328 "RF NA 0.0207492176709\n",
|
|
m@79
|
329 "KNN NA 0.221769305159\n",
|
|
m@79
|
330 "LDA NA 0.191217962613\n",
|
|
m@79
|
331 "SVM NA 0.250268813953\n",
|
|
m@82
|
332 "RF NA 0.117754824997\n",
|
|
m@79
|
333 "KNN NA 0.0814734970192\n",
|
|
m@79
|
334 "LDA NA 0.0839348156722\n",
|
|
m@79
|
335 "SVM NA 0.0881235182136\n",
|
|
m@82
|
336 "RF NA 0.0661199464142\n"
|
|
m@65
|
337 ]
|
|
m@65
|
338 }
|
|
m@65
|
339 ],
|
|
m@65
|
340 "source": [
|
|
m@79
|
341 "df_results = classification.classify_for_filenames(file_list=mapper.OUTPUT_FILES)"
|
|
m@65
|
342 ]
|
|
m@65
|
343 },
|
|
m@65
|
344 {
|
|
m@65
|
345 "cell_type": "code",
|
|
m@82
|
346 "execution_count": 5,
|
|
m@79
|
347 "metadata": {},
|
|
m@65
|
348 "outputs": [
|
|
m@65
|
349 {
|
|
m@65
|
350 "name": "stdout",
|
|
m@65
|
351 "output_type": "stream",
|
|
m@65
|
352 "text": [
|
|
m@79
|
353 "\\begin{tabular}{llrrrrr}\n",
|
|
m@79
|
354 "\\toprule\n",
|
|
m@79
|
355 " 0 & 1 & 2 & 3 & 4 & 5 & 6 \\\\\n",
|
|
m@79
|
356 "\\midrule\n",
|
|
m@79
|
357 " LDA & KNN & 0.151978 & 0.054739 & 0.023233 & 0.281734 & 0.085792 \\\\\n",
|
|
m@79
|
358 " LDA & LDA & 0.320670 & 0.150313 & 0.070247 & 0.198583 & 0.107355 \\\\\n",
|
|
m@79
|
359 " LDA & SVM & 0.023110 & 0.078763 & 0.050069 & 0.296356 & 0.089610 \\\\\n",
|
|
m@82
|
360 " LDA & RF & 0.075071 & 0.032517 & 0.022196 & 0.138929 & 0.052058 \\\\\n",
|
|
m@79
|
361 " PCA & KNN & 0.140644 & 0.052517 & 0.026873 & 0.220850 & 0.081427 \\\\\n",
|
|
m@79
|
362 " PCA & LDA & 0.175099 & 0.055029 & 0.045930 & 0.161503 & 0.083973 \\\\\n",
|
|
m@79
|
363 " PCA & SVM & 0.014927 & 0.047951 & 0.038673 & 0.245791 & 0.091864 \\\\\n",
|
|
m@82
|
364 " PCA & RF & 0.043688 & 0.029319 & 0.020051 & 0.109653 & 0.051269 \\\\\n",
|
|
m@79
|
365 " NMF & KNN & 0.114299 & 0.043058 & 0.028528 & 0.177820 & 0.079545 \\\\\n",
|
|
m@79
|
366 " NMF & LDA & 0.178244 & 0.058666 & 0.046366 & 0.166222 & 0.085643 \\\\\n",
|
|
m@79
|
367 " NMF & SVM & 0.016406 & 0.007813 & 0.007689 & 0.010789 & 0.011692 \\\\\n",
|
|
m@82
|
368 " NMF & RF & 0.060399 & 0.039489 & 0.024324 & 0.125063 & 0.041028 \\\\\n",
|
|
m@79
|
369 " SSNMF & KNN & 0.143227 & 0.043130 & 0.015224 & 0.227101 & 0.071541 \\\\\n",
|
|
m@79
|
370 " SSNMF & LDA & 0.183202 & 0.053345 & 0.038873 & 0.165382 & 0.081976 \\\\\n",
|
|
m@79
|
371 " SSNMF & SVM & 0.020578 & 0.010654 & 0.005361 & 0.018492 & 0.013882 \\\\\n",
|
|
m@82
|
372 " SSNMF & RF & 0.045948 & 0.026722 & 0.019617 & 0.092444 & 0.034744 \\\\\n",
|
|
m@79
|
373 " NA & KNN & 0.140075 & 0.051532 & 0.027336 & 0.221769 & 0.081473 \\\\\n",
|
|
m@79
|
374 " NA & LDA & 0.176954 & 0.059945 & 0.037882 & 0.191218 & 0.083935 \\\\\n",
|
|
m@79
|
375 " NA & SVM & 0.014949 & 0.046862 & 0.038291 & 0.250269 & 0.088124 \\\\\n",
|
|
m@82
|
376 " NA & RF & 0.068184 & 0.045879 & 0.020749 & 0.117755 & 0.066120 \\\\\n",
|
|
m@79
|
377 "\\bottomrule\n",
|
|
m@79
|
378 "\\end{tabular}\n",
|
|
m@79
|
379 "\n",
|
|
m@79
|
380 "\\begin{tabular}{llrrrrr}\n",
|
|
m@79
|
381 "\\toprule\n",
|
|
m@79
|
382 " 0 & 1 & 2 & 3 & 4 & 5 & 6 \\\\\n",
|
|
m@79
|
383 "\\midrule\n",
|
|
m@79
|
384 " LDA & LDA & 0.320670 & 0.150313 & 0.070247 & 0.198583 & 0.107355 \\\\\n",
|
|
m@79
|
385 " SSNMF & LDA & 0.183202 & 0.053345 & 0.038873 & 0.165382 & 0.081976 \\\\\n",
|
|
m@79
|
386 " NMF & LDA & 0.178244 & 0.058666 & 0.046366 & 0.166222 & 0.085643 \\\\\n",
|
|
m@79
|
387 " NA & LDA & 0.176954 & 0.059945 & 0.037882 & 0.191218 & 0.083935 \\\\\n",
|
|
m@79
|
388 " PCA & LDA & 0.175099 & 0.055029 & 0.045930 & 0.161503 & 0.083973 \\\\\n",
|
|
m@79
|
389 " LDA & KNN & 0.151978 & 0.054739 & 0.023233 & 0.281734 & 0.085792 \\\\\n",
|
|
m@79
|
390 " SSNMF & KNN & 0.143227 & 0.043130 & 0.015224 & 0.227101 & 0.071541 \\\\\n",
|
|
m@79
|
391 " PCA & KNN & 0.140644 & 0.052517 & 0.026873 & 0.220850 & 0.081427 \\\\\n",
|
|
m@79
|
392 " NA & KNN & 0.140075 & 0.051532 & 0.027336 & 0.221769 & 0.081473 \\\\\n",
|
|
m@79
|
393 " NMF & KNN & 0.114299 & 0.043058 & 0.028528 & 0.177820 & 0.079545 \\\\\n",
|
|
m@82
|
394 " LDA & RF & 0.075071 & 0.032517 & 0.022196 & 0.138929 & 0.052058 \\\\\n",
|
|
m@82
|
395 " NA & RF & 0.068184 & 0.045879 & 0.020749 & 0.117755 & 0.066120 \\\\\n",
|
|
m@82
|
396 " NMF & RF & 0.060399 & 0.039489 & 0.024324 & 0.125063 & 0.041028 \\\\\n",
|
|
m@82
|
397 " SSNMF & RF & 0.045948 & 0.026722 & 0.019617 & 0.092444 & 0.034744 \\\\\n",
|
|
m@82
|
398 " PCA & RF & 0.043688 & 0.029319 & 0.020051 & 0.109653 & 0.051269 \\\\\n",
|
|
m@79
|
399 " LDA & SVM & 0.023110 & 0.078763 & 0.050069 & 0.296356 & 0.089610 \\\\\n",
|
|
m@79
|
400 " SSNMF & SVM & 0.020578 & 0.010654 & 0.005361 & 0.018492 & 0.013882 \\\\\n",
|
|
m@79
|
401 " NMF & SVM & 0.016406 & 0.007813 & 0.007689 & 0.010789 & 0.011692 \\\\\n",
|
|
m@79
|
402 " NA & SVM & 0.014949 & 0.046862 & 0.038291 & 0.250269 & 0.088124 \\\\\n",
|
|
m@79
|
403 " PCA & SVM & 0.014927 & 0.047951 & 0.038673 & 0.245791 & 0.091864 \\\\\n",
|
|
m@79
|
404 "\\bottomrule\n",
|
|
m@79
|
405 "\\end{tabular}\n",
|
|
m@79
|
406 "\n"
|
|
m@65
|
407 ]
|
|
m@65
|
408 }
|
|
m@65
|
409 ],
|
|
m@65
|
410 "source": [
|
|
m@79
|
411 "print df_results.to_latex(index=False)\n",
|
|
m@79
|
412 "# sort by 'all'\n",
|
|
m@79
|
413 "df_results_sorted = df_results.sort_values(2, ascending=False, inplace=False)\n",
|
|
m@79
|
414 "df_results_sorted.head()\n",
|
|
m@79
|
415 "print df_results_sorted.to_latex(index=False)"
|
|
m@65
|
416 ]
|
|
m@65
|
417 },
|
|
m@65
|
418 {
|
|
m@65
|
419 "cell_type": "code",
|
|
m@65
|
420 "execution_count": 43,
|
|
m@79
|
421 "metadata": {},
|
|
m@65
|
422 "outputs": [
|
|
m@65
|
423 {
|
|
m@65
|
424 "name": "stdout",
|
|
m@65
|
425 "output_type": "stream",
|
|
m@65
|
426 "text": [
|
|
m@65
|
427 "detecting outliers...\n",
|
|
m@65
|
428 "most outliers \n",
|
|
m@65
|
429 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
430 "136 Botswana 0.611111 90 55\n",
|
|
m@65
|
431 "72 Ivory Coast 0.600000 15 9\n",
|
|
m@65
|
432 "95 Chad 0.545455 11 6\n",
|
|
m@65
|
433 "43 Benin 0.538462 26 14\n",
|
|
m@65
|
434 "86 Gambia 0.500000 50 25\n",
|
|
m@65
|
435 "20 Pakistan 0.494505 91 45\n",
|
|
m@65
|
436 "106 Nepal 0.473684 95 45\n",
|
|
m@65
|
437 "78 El Salvador 0.454545 33 15\n",
|
|
m@65
|
438 "64 Mozambique 0.441176 34 15\n",
|
|
m@65
|
439 "135 French Guiana 0.428571 28 12\n",
|
|
m@65
|
440 "least outliers \n",
|
|
m@65
|
441 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
442 "1 Lithuania 0.000000 47 0\n",
|
|
m@65
|
443 "119 Denmark 0.000000 16 0\n",
|
|
m@65
|
444 "27 South Korea 0.000000 11 0\n",
|
|
m@65
|
445 "120 Kazakhstan 0.011364 88 1\n",
|
|
m@65
|
446 "31 Czech Republic 0.024390 41 1\n",
|
|
m@65
|
447 "15 Netherlands 0.029851 67 2\n",
|
|
m@65
|
448 "30 Afghanistan 0.041667 24 1\n",
|
|
m@65
|
449 "105 Sudan 0.044118 68 3\n",
|
|
m@65
|
450 "102 Nicaragua 0.047619 21 1\n",
|
|
m@65
|
451 "0 Canada 0.050000 100 5\n"
|
|
m@65
|
452 ]
|
|
m@65
|
453 }
|
|
m@65
|
454 ],
|
|
m@65
|
455 "source": [
|
|
m@65
|
456 "# outliers\n",
|
|
m@65
|
457 "print \"detecting outliers...\"\n",
|
|
m@65
|
458 "#ddf = outliers.load_metadata(Yaudio, metadata_file=load_dataset.METADATA_FILE)\n",
|
|
m@65
|
459 "#X = np.concatenate(ldadata_list, axis=1)\n",
|
|
m@65
|
460 "X, Y, Yaudio = classification.load_data_from_pickle(mapper.OUTPUT_FILES[0])\n",
|
|
m@65
|
461 "df_global, threshold, MD = outliers.get_outliers_df(X, Y, chi2thr=0.999)\n",
|
|
m@65
|
462 "outliers.print_most_least_outliers_topN(df_global, N=10)"
|
|
m@65
|
463 ]
|
|
m@65
|
464 },
|
|
m@65
|
465 {
|
|
m@65
|
466 "cell_type": "code",
|
|
m@65
|
467 "execution_count": 46,
|
|
m@79
|
468 "metadata": {},
|
|
m@65
|
469 "outputs": [
|
|
m@65
|
470 {
|
|
m@65
|
471 "name": "stdout",
|
|
m@65
|
472 "output_type": "stream",
|
|
m@65
|
473 "text": [
|
|
m@65
|
474 "most outliers \n",
|
|
m@65
|
475 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
476 "43 Benin 0.500000 26 13\n",
|
|
m@65
|
477 "136 Botswana 0.488889 90 44\n",
|
|
m@65
|
478 "106 Nepal 0.421053 95 40\n",
|
|
m@65
|
479 "84 Belize 0.418605 43 18\n",
|
|
m@65
|
480 "19 Yemen 0.416667 12 5\n",
|
|
m@65
|
481 "least outliers \n",
|
|
m@65
|
482 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
483 "28 Tajikistan 0 19 0\n",
|
|
m@65
|
484 "119 Denmark 0 16 0\n",
|
|
m@65
|
485 "96 Uruguay 0 31 0\n",
|
|
m@65
|
486 "25 Republic of Serbia 0 16 0\n",
|
|
m@65
|
487 "27 South Korea 0 11 0\n",
|
|
m@65
|
488 "most outliers \n",
|
|
m@65
|
489 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
490 "117 Zimbabwe 0.533333 15 8\n",
|
|
m@65
|
491 "96 Uruguay 0.483871 31 15\n",
|
|
m@65
|
492 "68 Guinea 0.454545 11 5\n",
|
|
m@65
|
493 "63 Senegal 0.390244 41 16\n",
|
|
m@65
|
494 "86 Gambia 0.380000 50 19\n",
|
|
m@65
|
495 "least outliers \n",
|
|
m@65
|
496 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
497 "90 French Polynesia 0.000000 15 0\n",
|
|
m@65
|
498 "37 Rwanda 0.000000 17 0\n",
|
|
m@65
|
499 "119 Denmark 0.000000 16 0\n",
|
|
m@65
|
500 "18 New Zealand 0.000000 34 0\n",
|
|
m@65
|
501 "120 Kazakhstan 0.022727 88 2\n",
|
|
m@65
|
502 "most outliers \n",
|
|
m@65
|
503 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
504 "17 French Guiana 0.678571 28 19\n",
|
|
m@65
|
505 "136 Botswana 0.477778 90 43\n",
|
|
m@65
|
506 "72 Ivory Coast 0.400000 15 6\n",
|
|
m@65
|
507 "23 Azerbaijan 0.384615 13 5\n",
|
|
m@65
|
508 "106 Nepal 0.347368 95 33\n",
|
|
m@65
|
509 "least outliers \n",
|
|
m@65
|
510 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
511 "68 Guinea 0 11 0\n",
|
|
m@65
|
512 "55 Mali 0 17 0\n",
|
|
m@65
|
513 "77 Algeria 0 27 0\n",
|
|
m@65
|
514 "33 Saint Lucia 0 43 0\n",
|
|
m@65
|
515 "31 Czech Republic 0 41 0\n",
|
|
m@65
|
516 "most outliers \n",
|
|
m@65
|
517 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
518 "43 Benin 0.538462 26 14\n",
|
|
m@65
|
519 "20 Pakistan 0.461538 91 42\n",
|
|
m@65
|
520 "86 Gambia 0.360000 50 18\n",
|
|
m@65
|
521 "52 Indonesia 0.350000 100 35\n",
|
|
m@65
|
522 "136 Botswana 0.311111 90 28\n",
|
|
m@65
|
523 "least outliers \n",
|
|
m@65
|
524 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
525 "107 Kiribati 0 17 0\n",
|
|
m@65
|
526 "1 Lithuania 0 47 0\n",
|
|
m@65
|
527 "134 Paraguay 0 23 0\n",
|
|
m@65
|
528 "131 Tunisia 0 39 0\n",
|
|
m@65
|
529 "19 Yemen 0 12 0\n"
|
|
m@65
|
530 ]
|
|
m@65
|
531 }
|
|
m@65
|
532 ],
|
|
m@65
|
533 "source": [
|
|
m@65
|
534 "X_list, Y, Yaudio = pickle.load(open(mapper.OUTPUT_FILES[0],'rb'))\n",
|
|
m@65
|
535 "feat = X_list\n",
|
|
m@65
|
536 "feat_labels = ['rhy', 'mel', 'mfc', 'chr']\n",
|
|
m@65
|
537 "tabs_feat = []\n",
|
|
m@65
|
538 "for i in range(len(feat)):\n",
|
|
m@65
|
539 " XX = feat[i]\n",
|
|
m@65
|
540 " df_feat, threshold, MD = outliers.get_outliers_df(XX, Y, chi2thr=0.999)\n",
|
|
m@65
|
541 " outliers.print_most_least_outliers_topN(df_feat, N=5)"
|
|
m@65
|
542 ]
|
|
m@65
|
543 },
|
|
m@65
|
544 {
|
|
m@65
|
545 "cell_type": "code",
|
|
m@65
|
546 "execution_count": 48,
|
|
m@79
|
547 "metadata": {},
|
|
m@65
|
548 "outputs": [
|
|
m@65
|
549 {
|
|
m@65
|
550 "name": "stderr",
|
|
m@65
|
551 "output_type": "stream",
|
|
m@65
|
552 "text": [
|
|
m@65
|
553 "/homes/mp305/anaconda/lib/python2.7/site-packages/IPython/core/interactiveshell.py:2882: DtypeWarning: Columns (0,1,2,4,5,6,7,8,10,11,12,13,14,15,16,17,19,21,22,23,24,25,26,27,29,31,35,38,39,40,41,44,45,48,55,56,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,93,95,96) have mixed types. Specify dtype option on import or set low_memory=False.\n",
|
|
m@65
|
554 " exec(code_obj, self.user_global_ns, self.user_ns)\n",
|
|
m@65
|
555 "/homes/mp305/anaconda/lib/python2.7/site-packages/pysal/weights/weights.py:189: UserWarning: There are 21 disconnected observations\n",
|
|
m@65
|
556 " warnings.warn(\"There are %d disconnected observations\" % ni)\n",
|
|
m@65
|
557 "/homes/mp305/anaconda/lib/python2.7/site-packages/pysal/weights/weights.py:190: UserWarning: Island ids: 3, 6, 26, 35, 39, 45, 52, 61, 62, 66, 77, 85, 94, 97, 98, 102, 103, 107, 110, 120, 121\n",
|
|
m@65
|
558 " warnings.warn(\"Island ids: %s\" % ', '.join(str(island) for island in self.islands))\n"
|
|
m@65
|
559 ]
|
|
m@65
|
560 },
|
|
m@65
|
561 {
|
|
m@65
|
562 "name": "stdout",
|
|
m@65
|
563 "output_type": "stream",
|
|
m@65
|
564 "text": [
|
|
m@65
|
565 "Antigua and Barbuda\n",
|
|
m@65
|
566 "Australia\n",
|
|
m@65
|
567 "Cuba\n",
|
|
m@65
|
568 "Fiji\n",
|
|
m@65
|
569 "French Polynesia\n",
|
|
m@65
|
570 "Grenada\n",
|
|
m@65
|
571 "Iceland\n",
|
|
m@65
|
572 "Jamaica\n",
|
|
m@65
|
573 "Japan\n",
|
|
m@65
|
574 "Kiribati\n",
|
|
m@65
|
575 "Malta\n",
|
|
m@65
|
576 "New Zealand\n",
|
|
m@65
|
577 "Philippines\n",
|
|
m@65
|
578 "Puerto Rico\n",
|
|
m@65
|
579 "Republic of Serbia\n",
|
|
m@65
|
580 "Saint Lucia\n",
|
|
m@65
|
581 "Samoa\n",
|
|
m@65
|
582 "Solomon Islands\n",
|
|
m@65
|
583 "South Korea\n",
|
|
m@65
|
584 "The Bahamas\n",
|
|
m@65
|
585 "Trinidad and Tobago\n",
|
|
m@65
|
586 "328\n",
|
|
m@65
|
587 "210\n",
|
|
m@65
|
588 "194\n",
|
|
m@65
|
589 "85\n",
|
|
m@65
|
590 "388\n",
|
|
m@65
|
591 "266\n",
|
|
m@65
|
592 "309\n",
|
|
m@65
|
593 "455\n",
|
|
m@65
|
594 "365\n",
|
|
m@65
|
595 "282\n",
|
|
m@65
|
596 "197\n",
|
|
m@65
|
597 "122\n",
|
|
m@65
|
598 "206\n",
|
|
m@65
|
599 "457\n",
|
|
m@65
|
600 "298\n",
|
|
m@65
|
601 "597\n",
|
|
m@65
|
602 "354\n",
|
|
m@65
|
603 "191\n",
|
|
m@65
|
604 "193\n",
|
|
m@65
|
605 "198\n",
|
|
m@65
|
606 "263\n",
|
|
m@65
|
607 "334\n",
|
|
m@65
|
608 "812\n",
|
|
m@65
|
609 "415\n",
|
|
m@65
|
610 "44\n",
|
|
m@65
|
611 "107\n",
|
|
m@65
|
612 "366\n",
|
|
m@65
|
613 "323\n",
|
|
m@65
|
614 "450\n",
|
|
m@65
|
615 "116\n",
|
|
m@65
|
616 "150\n",
|
|
m@65
|
617 "260\n",
|
|
m@65
|
618 "230\n",
|
|
m@65
|
619 "118\n",
|
|
m@65
|
620 "389\n",
|
|
m@65
|
621 "237\n",
|
|
m@65
|
622 "274\n",
|
|
m@65
|
623 "466\n",
|
|
m@65
|
624 "147\n",
|
|
m@65
|
625 "134\n",
|
|
m@65
|
626 "86\n",
|
|
m@65
|
627 "91\n",
|
|
m@65
|
628 "574\n",
|
|
m@65
|
629 "111\n",
|
|
m@65
|
630 "296\n",
|
|
m@65
|
631 "221\n",
|
|
m@65
|
632 "261\n",
|
|
m@65
|
633 "224\n",
|
|
m@65
|
634 "190\n",
|
|
m@65
|
635 "150\n",
|
|
m@65
|
636 "139\n",
|
|
m@65
|
637 "350\n",
|
|
m@65
|
638 "268\n",
|
|
m@65
|
639 "453\n",
|
|
m@65
|
640 "192\n",
|
|
m@65
|
641 "468\n",
|
|
m@65
|
642 "266\n",
|
|
m@65
|
643 "187\n",
|
|
m@65
|
644 "275\n",
|
|
m@65
|
645 "337\n",
|
|
m@65
|
646 "179\n",
|
|
m@65
|
647 "366\n",
|
|
m@65
|
648 "211\n",
|
|
m@65
|
649 "213\n",
|
|
m@65
|
650 "428\n",
|
|
m@65
|
651 "468\n",
|
|
m@65
|
652 "164\n",
|
|
m@65
|
653 "348\n",
|
|
m@65
|
654 "328\n",
|
|
m@65
|
655 "193\n",
|
|
m@65
|
656 "197\n",
|
|
m@65
|
657 "193\n",
|
|
m@65
|
658 "166\n",
|
|
m@65
|
659 "290\n",
|
|
m@65
|
660 "196\n",
|
|
m@65
|
661 "224\n",
|
|
m@65
|
662 "111\n",
|
|
m@65
|
663 "258\n",
|
|
m@65
|
664 "295\n",
|
|
m@65
|
665 "227\n",
|
|
m@65
|
666 "252\n",
|
|
m@65
|
667 "433\n",
|
|
m@65
|
668 "305\n",
|
|
m@65
|
669 "290\n",
|
|
m@65
|
670 "183\n",
|
|
m@65
|
671 "243\n",
|
|
m@65
|
672 "63\n",
|
|
m@65
|
673 "197\n",
|
|
m@65
|
674 "274\n",
|
|
m@65
|
675 "363\n",
|
|
m@65
|
676 "113\n",
|
|
m@65
|
677 "192\n",
|
|
m@65
|
678 "258\n",
|
|
m@65
|
679 "494\n",
|
|
m@65
|
680 "299\n",
|
|
m@65
|
681 "484\n",
|
|
m@65
|
682 "198\n",
|
|
m@65
|
683 "191\n",
|
|
m@65
|
684 "174\n",
|
|
m@65
|
685 "280\n",
|
|
m@65
|
686 "735\n",
|
|
m@65
|
687 "211\n",
|
|
m@65
|
688 "221\n",
|
|
m@65
|
689 "134\n",
|
|
m@65
|
690 "125\n",
|
|
m@65
|
691 "119\n",
|
|
m@65
|
692 "151\n",
|
|
m@65
|
693 "203\n",
|
|
m@65
|
694 "229\n",
|
|
m@65
|
695 "430\n",
|
|
m@65
|
696 "311\n",
|
|
m@65
|
697 "424\n",
|
|
m@65
|
698 "337\n",
|
|
m@65
|
699 "268\n",
|
|
m@65
|
700 "175\n",
|
|
m@65
|
701 "228\n",
|
|
m@65
|
702 "175\n",
|
|
m@65
|
703 "437\n",
|
|
m@65
|
704 "284\n",
|
|
m@65
|
705 "129\n",
|
|
m@65
|
706 "366\n",
|
|
m@65
|
707 "222\n",
|
|
m@65
|
708 "66\n",
|
|
m@65
|
709 "498\n",
|
|
m@65
|
710 "400\n",
|
|
m@65
|
711 "430\n",
|
|
m@65
|
712 "187\n",
|
|
m@65
|
713 "470\n",
|
|
m@65
|
714 "298\n",
|
|
m@65
|
715 "231\n",
|
|
m@65
|
716 "272\n",
|
|
m@65
|
717 "261\n",
|
|
m@65
|
718 "239\n",
|
|
m@65
|
719 "154\n",
|
|
m@65
|
720 "22\n",
|
|
m@65
|
721 "426\n",
|
|
m@65
|
722 "332\n",
|
|
m@65
|
723 "most outliers \n",
|
|
m@65
|
724 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
725 "46 China 0.260000 100 26\n",
|
|
m@65
|
726 "67 Brazil 0.240000 100 24\n",
|
|
m@65
|
727 "101 Colombia 0.211111 90 19\n",
|
|
m@65
|
728 "64 Mozambique 0.205882 34 7\n",
|
|
m@65
|
729 "76 Iran 0.188679 53 10\n",
|
|
m@65
|
730 "65 Uganda 0.176471 85 15\n",
|
|
m@65
|
731 "27 Kenya 0.164948 97 16\n",
|
|
m@65
|
732 "126 South Sudan 0.163043 92 15\n",
|
|
m@65
|
733 "24 Azerbaijan 0.153846 13 2\n",
|
|
m@65
|
734 "23 India 0.147368 95 14\n",
|
|
m@65
|
735 "least outliers \n",
|
|
m@65
|
736 " Country Outliers N_Country N_Outliers\n",
|
|
m@65
|
737 "0 Canada 0 100 0\n",
|
|
m@65
|
738 "95 Portugal 0 100 0\n",
|
|
m@65
|
739 "94 Iraq 0 87 0\n",
|
|
m@65
|
740 "93 Grenada 0 37 0\n",
|
|
m@65
|
741 "90 French Polynesia 0 15 0\n",
|
|
m@65
|
742 "89 Croatia 0 31 0\n",
|
|
m@65
|
743 "88 Morocco 0 40 0\n",
|
|
m@65
|
744 "87 Philippines 0 100 0\n",
|
|
m@65
|
745 "86 Gambia 0 50 0\n",
|
|
m@65
|
746 "85 Sierra Leone 0 100 0\n"
|
|
m@65
|
747 ]
|
|
m@65
|
748 }
|
|
m@65
|
749 ],
|
|
m@65
|
750 "source": [
|
|
m@65
|
751 "dataset, ddf, w_dict = outliers.load_data(mapper.OUTPUT_FILES[0], '../data/metadata_BLSM_language_all.csv')\n",
|
|
m@65
|
752 "df_local = outliers.get_local_outliers_df(X, Y, w_dict)\n",
|
|
m@65
|
753 "outliers.print_most_least_outliers_topN(df_local, N=10)"
|
|
m@65
|
754 ]
|
|
m@65
|
755 },
|
|
m@65
|
756 {
|
|
m@65
|
757 "cell_type": "code",
|
|
m@65
|
758 "execution_count": 53,
|
|
m@79
|
759 "metadata": {},
|
|
m@65
|
760 "outputs": [
|
|
m@65
|
761 {
|
|
m@65
|
762 "name": "stdout",
|
|
m@65
|
763 "output_type": "stream",
|
|
m@65
|
764 "text": [
|
|
m@65
|
765 "\\begin{tabular}{llll}\n",
|
|
m@65
|
766 "\\toprule\n",
|
|
m@65
|
767 "{} & 0 & 1 & 2 \\\\\n",
|
|
m@65
|
768 "\\midrule\n",
|
|
m@65
|
769 "0 & (Uruguay, 19) & (Switzerland, 21) & (Austria, 30) \\\\\n",
|
|
m@65
|
770 "1 & (Nigeria, 17) & (United Republic of Tanzania, 20) & (Uganda, 29) \\\\\n",
|
|
m@65
|
771 "2 & (Greece, 27) & (Armenia, 32) & (Ukraine, 45) \\\\\n",
|
|
m@65
|
772 "3 & (Russia, 27) & (Israel, 29) & (Kazakhstan, 45) \\\\\n",
|
|
m@65
|
773 "4 & (Uzbekistan, 1) & (Democratic Republic of the Congo, 4) & (Botswana, 5) \\\\\n",
|
|
m@65
|
774 "5 & (Swaziland, 1) & (Senegal, 4) & (Gambia, 5) \\\\\n",
|
|
m@65
|
775 "6 & (Papua New Guinea, 9) & (Colombia, 12) & (French Guiana, 16) \\\\\n",
|
|
m@65
|
776 "7 & (Western Sahara, 13) & (Pakistan, 14) & (Iraq, 33) \\\\\n",
|
|
m@65
|
777 "8 & (Canada, 47) & (Norway, 47) & (South Sudan, 49) \\\\\n",
|
|
m@65
|
778 "9 & (Ghana, 3) & (Zambia, 8) & None \\\\\n",
|
|
m@65
|
779 "10 & (Sierra Leone, 18) & (Haiti, 19) & (Trinidad and Tobago, 22) \\\\\n",
|
|
m@65
|
780 "11 & (Portugal, 15) & (Netherlands, 17) & (Austria, 19) \\\\\n",
|
|
m@65
|
781 "12 & (Swaziland, 24) & (South Sudan, 38) & (Lesotho, 41) \\\\\n",
|
|
m@65
|
782 "13 & (Vietnam, 20) & (Iraq, 27) & (Nepal, 33) \\\\\n",
|
|
m@65
|
783 "14 & (Jamaica, 20) & (Cuba, 21) & (Zambia, 22) \\\\\n",
|
|
m@65
|
784 "15 & (Chile, 30) & (Kazakhstan, 31) & (Trinidad and Tobago, 35) \\\\\n",
|
|
m@65
|
785 "16 & (Philippines, 6) & (Ghana, 8) & (Botswana, 12) \\\\\n",
|
|
m@65
|
786 "17 & (Nigeria, 37) & (Australia, 42) & (Solomon Islands, 47) \\\\\n",
|
|
m@65
|
787 "18 & (Senegal, 1) & None & None \\\\\n",
|
|
m@65
|
788 "19 & (Kyrgyzstan, 22) & (Pakistan, 24) & (Turkey, 37) \\\\\n",
|
|
m@65
|
789 "\\bottomrule\n",
|
|
m@65
|
790 "\\end{tabular}\n",
|
|
m@65
|
791 "\n"
|
|
m@65
|
792 ]
|
|
m@65
|
793 }
|
|
m@65
|
794 ],
|
|
m@65
|
795 "source": [
|
|
m@65
|
796 "centroids, cl_pred = outliers.get_country_clusters(X, bestncl=20)\n",
|
|
m@65
|
797 "ddf['Clusters'] = cl_pred\n",
|
|
m@65
|
798 "outliers.print_clusters_metadata(ddf, cl_pred)"
|
|
m@65
|
799 ]
|
|
m@65
|
800 },
|
|
m@65
|
801 {
|
|
m@65
|
802 "cell_type": "code",
|
|
m@65
|
803 "execution_count": 54,
|
|
m@65
|
804 "metadata": {
|
|
m@65
|
805 "collapsed": true
|
|
m@65
|
806 },
|
|
m@65
|
807 "outputs": [],
|
|
m@65
|
808 "source": [
|
|
m@65
|
809 "cluster_freq = utils.get_cluster_freq_linear(X, Y, centroids)"
|
|
m@65
|
810 ]
|
|
m@65
|
811 },
|
|
m@65
|
812 {
|
|
m@65
|
813 "cell_type": "code",
|
|
m@65
|
814 "execution_count": 55,
|
|
m@79
|
815 "metadata": {},
|
|
m@65
|
816 "outputs": [
|
|
m@65
|
817 {
|
|
m@65
|
818 "data": {
|
|
m@65
|
819 "text/plain": [
|
|
m@65
|
820 "(137, 20)"
|
|
m@65
|
821 ]
|
|
m@65
|
822 },
|
|
m@65
|
823 "execution_count": 55,
|
|
m@65
|
824 "metadata": {},
|
|
m@65
|
825 "output_type": "execute_result"
|
|
m@65
|
826 }
|
|
m@65
|
827 ],
|
|
m@65
|
828 "source": [
|
|
m@65
|
829 "cluster_freq.shape"
|
|
m@65
|
830 ]
|
|
m@65
|
831 },
|
|
m@65
|
832 {
|
|
m@65
|
833 "cell_type": "code",
|
|
m@65
|
834 "execution_count": 62,
|
|
m@79
|
835 "metadata": {},
|
|
m@65
|
836 "outputs": [
|
|
m@65
|
837 {
|
|
m@65
|
838 "data": {
|
|
m@65
|
839 "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgoAAAD/CAYAAACQAf7WAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIiIiFb8H1TxBwEWfqzuAAAAAElFTkSuQmCC\n",
|
|
m@65
|
840 "text/plain": [
|
|
m@65
|
841 "<matplotlib.figure.Figure at 0x7f74fa288650>"
|
|
m@65
|
842 ]
|
|
m@65
|
843 },
|
|
m@65
|
844 "metadata": {},
|
|
m@65
|
845 "output_type": "display_data"
|
|
m@65
|
846 }
|
|
m@65
|
847 ],
|
|
m@65
|
848 "source": [
|
|
m@65
|
849 "plt.figure()\n",
|
|
m@65
|
850 "plt.imshow(cluster_freq, aspect='auto')\n",
|
|
m@65
|
851 "plt.yticks(np.arange(len(cluster_freq)), np.unique(Y));"
|
|
m@65
|
852 ]
|
|
m@65
|
853 },
|
|
m@65
|
854 {
|
|
m@65
|
855 "cell_type": "code",
|
|
m@65
|
856 "execution_count": null,
|
|
m@65
|
857 "metadata": {
|
|
m@65
|
858 "collapsed": true
|
|
m@65
|
859 },
|
|
m@65
|
860 "outputs": [],
|
|
m@65
|
861 "source": []
|
|
m@65
|
862 }
|
|
m@65
|
863 ],
|
|
m@65
|
864 "metadata": {
|
|
m@65
|
865 "kernelspec": {
|
|
m@65
|
866 "display_name": "Python 2",
|
|
m@65
|
867 "language": "python",
|
|
m@65
|
868 "name": "python2"
|
|
m@65
|
869 },
|
|
m@65
|
870 "language_info": {
|
|
m@65
|
871 "codemirror_mode": {
|
|
m@65
|
872 "name": "ipython",
|
|
m@65
|
873 "version": 2
|
|
m@65
|
874 },
|
|
m@65
|
875 "file_extension": ".py",
|
|
m@65
|
876 "mimetype": "text/x-python",
|
|
m@65
|
877 "name": "python",
|
|
m@65
|
878 "nbconvert_exporter": "python",
|
|
m@65
|
879 "pygments_lexer": "ipython2",
|
|
m@65
|
880 "version": "2.7.12"
|
|
m@65
|
881 }
|
|
m@65
|
882 },
|
|
m@65
|
883 "nbformat": 4,
|
|
m@65
|
884 "nbformat_minor": 2
|
|
m@65
|
885 }
|
