Commit | Line | Data |
---|---|---|
d444b27a BA |
1 | { |
2 | "cells": [ | |
3 | { | |
4 | "cell_type": "markdown", | |
5 | "metadata": {}, | |
6 | "source": [ | |
7 | "\n", | |
8 | "\n", | |
9 | "<h2>Introduction</h2>\n", | |
10 | "\n", | |
11 | "J'ai fait quelques essais dans différentes configurations pour la méthode \"Neighbors\"\n", | |
12 | "(la seule dont on a parlé).<br>Il semble que le mieux soit\n", | |
13 | "\n", | |
14 | " * simtype=\"exo\" ou \"mix\" : similarités exogènes avec/sans endogènes (fenêtre optimisée par VC)\n", | |
15 | " * same_season=FALSE : les indices pour la validation croisée ne tiennent pas compte des saisons\n", | |
16 | " * mix_strategy=\"mult\" : on multiplie les poids (au lieu d'en éteindre)\n", | |
17 | "\n", | |
18 | "J'ai systématiquement comparé à une approche naïve : la moyennes des lendemains des jours\n", | |
19 | "\"similaires\" dans tout le passé ; à chaque fois sans prédiction du saut (sauf pour Neighbors :\n", | |
20 | "prédiction basée sur les poids calculés).\n", | |
21 | "\n", | |
22 | "Ensuite j'affiche les erreurs, quelques courbes prévues/mesurées, quelques filaments puis les\n", | |
23 | "histogrammes de quelques poids. Concernant les graphes de filaments, la moitié gauche du graphe\n", | |
24 | "correspond aux jours similaires au jour courant, tandis que la moitié droite affiche les\n", | |
25 | "lendemains : ce sont donc les voisinages tels qu'utilisés dans l'algorithme.\n", | |
26 | "\n" | |
27 | ] | |
28 | }, | |
29 | { | |
30 | "cell_type": "code", | |
31 | "execution_count": null, | |
32 | "metadata": { | |
33 | "collapsed": false | |
34 | }, | |
35 | "outputs": [], | |
36 | "source": [ | |
37 | "library(talweg)\n", | |
38 | "\n", | |
39 | "ts_data = read.csv(system.file(\"extdata\",\"pm10_mesures_H_loc_report.csv\",package=\"talweg\"))\n", | |
40 | "exo_data = read.csv(system.file(\"extdata\",\"meteo_extra_noNAs.csv\",package=\"talweg\"))\n", | |
41 | "data = getData(ts_data, exo_data, input_tz = \"Europe/Paris\", working_tz=\"Europe/Paris\",\n", | |
42 | "\tpredict_at=7) #predict from P+1 to P+H included\n", | |
43 | "\n", | |
44 | "indices_ch = seq(as.Date(\"2015-01-18\"),as.Date(\"2015-01-24\"),\"days\")\n", | |
45 | "indices_ep = seq(as.Date(\"2015-03-15\"),as.Date(\"2015-03-21\"),\"days\")\n", | |
46 | "indices_np = seq(as.Date(\"2015-04-26\"),as.Date(\"2015-05-02\"),\"days\")" | |
47 | ] | |
48 | }, | |
49 | { | |
50 | "cell_type": "markdown", | |
51 | "metadata": {}, | |
52 | "source": [ | |
53 | "\n", | |
54 | "\n", | |
55 | "<h2 style=\"color:blue;font-size:2em\">Pollution par chauffage</h2>" | |
56 | ] | |
57 | }, | |
58 | { | |
59 | "cell_type": "code", | |
60 | "execution_count": null, | |
61 | "metadata": { | |
62 | "collapsed": false | |
63 | }, | |
64 | "outputs": [], | |
65 | "source": [ | |
66 | "p_nn_exo = computeForecast(data, indices_ch, \"Neighbors\", \"Neighbors\",\n", | |
67 | "\thorizon=3, simtype=\"exo\")\n", | |
68 | "p_nn_mix = computeForecast(data, indices_ch, \"Neighbors\", \"Neighbors\",\n", | |
69 | "\thorizon=3, simtype=\"mix\")\n", | |
70 | "p_az = computeForecast(data, indices_ch, \"Average\", \"Zero\",\n", | |
71 | "\thorizon=3)\n", | |
72 | "p_pz = computeForecast(data, indices_ch, \"Persistence\", \"Zero\",\n", | |
73 | "\thorizon=3, same_day=TRUE)" | |
74 | ] | |
75 | }, | |
76 | { | |
77 | "cell_type": "code", | |
78 | "execution_count": null, | |
79 | "metadata": { | |
80 | "collapsed": false | |
81 | }, | |
82 | "outputs": [], | |
83 | "source": [ | |
84 | "e_nn_exo = computeError(data, p_nn_exo, 3)\n", | |
85 | "e_nn_mix = computeError(data, p_nn_mix, 3)\n", | |
86 | "e_az = computeError(data, p_az, 3)\n", | |
87 | "e_pz = computeError(data, p_pz, 3)\n", | |
88 | "options(repr.plot.width=9, repr.plot.height=7)\n", | |
89 | "plotError(list(e_nn_mix, e_pz, e_az, e_nn_exo), cols=c(1,2,colors()[258], 4))\n", | |
90 | "\n", | |
91 | "# Noir: neighbors_mix, bleu: neighbors_exo, vert: moyenne, rouge: persistence\n", | |
92 | "\n", | |
93 | "i_np = which.min(e_nn_exo$abs$indices)\n", | |
94 | "i_p = which.max(e_nn_exo$abs$indices)" | |
95 | ] | |
96 | }, | |
97 | { | |
98 | "cell_type": "code", | |
99 | "execution_count": null, | |
100 | "metadata": { | |
101 | "collapsed": false | |
102 | }, | |
103 | "outputs": [], | |
104 | "source": [ | |
105 | "options(repr.plot.width=9, repr.plot.height=4)\n", | |
106 | "par(mfrow=c(1,2))\n", | |
107 | "\n", | |
108 | "plotPredReal(data, p_nn_exo, i_np); title(paste(\"PredReal nn exo day\",i_np))\n", | |
109 | "plotPredReal(data, p_nn_exo, i_p); title(paste(\"PredReal nn exo day\",i_p))\n", | |
110 | "\n", | |
111 | "plotPredReal(data, p_nn_mix, i_np); title(paste(\"PredReal nn mix day\",i_np))\n", | |
112 | "plotPredReal(data, p_nn_mix, i_p); title(paste(\"PredReal nn mix day\",i_p))\n", | |
113 | "\n", | |
114 | "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", | |
115 | "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", | |
116 | "\n", | |
117 | "# Bleu: prévue, noir: réalisée" | |
118 | ] | |
119 | }, | |
120 | { | |
121 | "cell_type": "code", | |
122 | "execution_count": null, | |
123 | "metadata": { | |
124 | "collapsed": false | |
125 | }, | |
126 | "outputs": [], | |
127 | "source": [ | |
128 | "par(mfrow=c(1,2))\n", | |
129 | "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", | |
130 | "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", | |
131 | "\n", | |
132 | "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", | |
133 | "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" | |
134 | ] | |
135 | }, | |
136 | { | |
137 | "cell_type": "code", | |
138 | "execution_count": null, | |
139 | "metadata": { | |
140 | "collapsed": false | |
141 | }, | |
142 | "outputs": [], | |
143 | "source": [ | |
144 | "par(mfrow=c(1,2))\n", | |
145 | "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", | |
146 | "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", | |
147 | "\n", | |
148 | "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", | |
149 | "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" | |
150 | ] | |
151 | }, | |
152 | { | |
153 | "cell_type": "code", | |
154 | "execution_count": null, | |
155 | "metadata": { | |
156 | "collapsed": false | |
157 | }, | |
158 | "outputs": [], | |
159 | "source": [ | |
160 | "par(mfrow=c(1,2))\n", | |
161 | "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", | |
162 | "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", | |
163 | "\n", | |
164 | "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", | |
165 | "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", | |
166 | "\n", | |
167 | "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" | |
168 | ] | |
169 | }, | |
170 | { | |
171 | "cell_type": "code", | |
172 | "execution_count": null, | |
173 | "metadata": { | |
174 | "collapsed": false | |
175 | }, | |
176 | "outputs": [], | |
177 | "source": [ | |
178 | "par(mfrow=c(1,2))\n", | |
179 | "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", | |
180 | "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", | |
181 | "\n", | |
182 | "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", | |
183 | "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", | |
184 | "\n", | |
185 | "# - pollué à gauche, + pollué à droite" | |
186 | ] | |
187 | }, | |
188 | { | |
189 | "cell_type": "code", | |
190 | "execution_count": null, | |
191 | "metadata": { | |
192 | "collapsed": false | |
193 | }, | |
194 | "outputs": [], | |
195 | "source": [ | |
196 | "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", | |
197 | "p_nn_exo$getParams(i_np)$window\n", | |
198 | "p_nn_exo$getParams(i_p)$window\n", | |
199 | "\n", | |
200 | "p_nn_mix$getParams(i_np)$window\n", | |
201 | "p_nn_mix$getParams(i_p)$window" | |
202 | ] | |
203 | }, | |
204 | { | |
205 | "cell_type": "markdown", | |
206 | "metadata": {}, | |
207 | "source": [ | |
208 | "\n", | |
209 | "\n", | |
210 | "<h2 style=\"color:blue;font-size:2em\">Pollution par épandage</h2>" | |
211 | ] | |
212 | }, | |
213 | { | |
214 | "cell_type": "code", | |
215 | "execution_count": null, | |
216 | "metadata": { | |
217 | "collapsed": false | |
218 | }, | |
219 | "outputs": [], | |
220 | "source": [ | |
221 | "p_nn_exo = computeForecast(data, indices_ep, \"Neighbors\", \"Neighbors\",\n", | |
222 | "\thorizon=3, simtype=\"exo\")\n", | |
223 | "p_nn_mix = computeForecast(data, indices_ep, \"Neighbors\", \"Neighbors\",\n", | |
224 | "\thorizon=3, simtype=\"mix\")\n", | |
225 | "p_az = computeForecast(data, indices_ep, \"Average\", \"Zero\",\n", | |
226 | "\thorizon=3)\n", | |
227 | "p_pz = computeForecast(data, indices_ep, \"Persistence\", \"Zero\",\n", | |
228 | "\thorizon=3, same_day=TRUE)" | |
229 | ] | |
230 | }, | |
231 | { | |
232 | "cell_type": "code", | |
233 | "execution_count": null, | |
234 | "metadata": { | |
235 | "collapsed": false | |
236 | }, | |
237 | "outputs": [], | |
238 | "source": [ | |
239 | "e_nn_exo = computeError(data, p_nn_exo, 3)\n", | |
240 | "e_nn_mix = computeError(data, p_nn_mix, 3)\n", | |
241 | "e_az = computeError(data, p_az, 3)\n", | |
242 | "e_pz = computeError(data, p_pz, 3)\n", | |
243 | "options(repr.plot.width=9, repr.plot.height=7)\n", | |
244 | "plotError(list(e_nn_mix, e_pz, e_az, e_nn_exo), cols=c(1,2,colors()[258], 4))\n", | |
245 | "\n", | |
246 | "# Noir: neighbors_mix, bleu: neighbors_exo, vert: moyenne, rouge: persistence\n", | |
247 | "\n", | |
248 | "i_np = which.min(e_nn_exo$abs$indices)\n", | |
249 | "i_p = which.max(e_nn_exo$abs$indices)" | |
250 | ] | |
251 | }, | |
252 | { | |
253 | "cell_type": "code", | |
254 | "execution_count": null, | |
255 | "metadata": { | |
256 | "collapsed": false | |
257 | }, | |
258 | "outputs": [], | |
259 | "source": [ | |
260 | "options(repr.plot.width=9, repr.plot.height=4)\n", | |
261 | "par(mfrow=c(1,2))\n", | |
262 | "\n", | |
263 | "plotPredReal(data, p_nn_exo, i_np); title(paste(\"PredReal nn exo day\",i_np))\n", | |
264 | "plotPredReal(data, p_nn_exo, i_p); title(paste(\"PredReal nn exo day\",i_p))\n", | |
265 | "\n", | |
266 | "plotPredReal(data, p_nn_mix, i_np); title(paste(\"PredReal nn mix day\",i_np))\n", | |
267 | "plotPredReal(data, p_nn_mix, i_p); title(paste(\"PredReal nn mix day\",i_p))\n", | |
268 | "\n", | |
269 | "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", | |
270 | "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", | |
271 | "\n", | |
272 | "# Bleu: prévue, noir: réalisée" | |
273 | ] | |
274 | }, | |
275 | { | |
276 | "cell_type": "code", | |
277 | "execution_count": null, | |
278 | "metadata": { | |
279 | "collapsed": false | |
280 | }, | |
281 | "outputs": [], | |
282 | "source": [ | |
283 | "par(mfrow=c(1,2))\n", | |
284 | "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", | |
285 | "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", | |
286 | "\n", | |
287 | "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", | |
288 | "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" | |
289 | ] | |
290 | }, | |
291 | { | |
292 | "cell_type": "code", | |
293 | "execution_count": null, | |
294 | "metadata": { | |
295 | "collapsed": false | |
296 | }, | |
297 | "outputs": [], | |
298 | "source": [ | |
299 | "par(mfrow=c(1,2))\n", | |
300 | "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", | |
301 | "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", | |
302 | "\n", | |
303 | "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", | |
304 | "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" | |
305 | ] | |
306 | }, | |
307 | { | |
308 | "cell_type": "code", | |
309 | "execution_count": null, | |
310 | "metadata": { | |
311 | "collapsed": false | |
312 | }, | |
313 | "outputs": [], | |
314 | "source": [ | |
315 | "par(mfrow=c(1,2))\n", | |
316 | "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", | |
317 | "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", | |
318 | "\n", | |
319 | "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", | |
320 | "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", | |
321 | "\n", | |
322 | "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" | |
323 | ] | |
324 | }, | |
325 | { | |
326 | "cell_type": "code", | |
327 | "execution_count": null, | |
328 | "metadata": { | |
329 | "collapsed": false | |
330 | }, | |
331 | "outputs": [], | |
332 | "source": [ | |
333 | "par(mfrow=c(1,2))\n", | |
334 | "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", | |
335 | "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", | |
336 | "\n", | |
337 | "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", | |
338 | "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", | |
339 | "\n", | |
340 | "# - pollué à gauche, + pollué à droite" | |
341 | ] | |
342 | }, | |
343 | { | |
344 | "cell_type": "code", | |
345 | "execution_count": null, | |
346 | "metadata": { | |
347 | "collapsed": false | |
348 | }, | |
349 | "outputs": [], | |
350 | "source": [ | |
351 | "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", | |
352 | "p_nn_exo$getParams(i_np)$window\n", | |
353 | "p_nn_exo$getParams(i_p)$window\n", | |
354 | "\n", | |
355 | "p_nn_mix$getParams(i_np)$window\n", | |
356 | "p_nn_mix$getParams(i_p)$window" | |
357 | ] | |
358 | }, | |
359 | { | |
360 | "cell_type": "markdown", | |
361 | "metadata": {}, | |
362 | "source": [ | |
363 | "\n", | |
364 | "\n", | |
365 | "<h2 style=\"color:blue;font-size:2em\">Semaine non polluée</h2>" | |
366 | ] | |
367 | }, | |
368 | { | |
369 | "cell_type": "code", | |
370 | "execution_count": null, | |
371 | "metadata": { | |
372 | "collapsed": false | |
373 | }, | |
374 | "outputs": [], | |
375 | "source": [ | |
376 | "p_nn_exo = computeForecast(data, indices_np, \"Neighbors\", \"Neighbors\",\n", | |
377 | "\thorizon=3, simtype=\"exo\")\n", | |
378 | "p_nn_mix = computeForecast(data, indices_np, \"Neighbors\", \"Neighbors\",\n", | |
379 | "\thorizon=3, simtype=\"mix\")\n", | |
380 | "p_az = computeForecast(data, indices_np, \"Average\", \"Zero\",\n", | |
381 | "\thorizon=3)\n", | |
382 | "p_pz = computeForecast(data, indices_np, \"Persistence\", \"Zero\",\n", | |
383 | "\thorizon=3, same_day=TRUE)" | |
384 | ] | |
385 | }, | |
386 | { | |
387 | "cell_type": "code", | |
388 | "execution_count": null, | |
389 | "metadata": { | |
390 | "collapsed": false | |
391 | }, | |
392 | "outputs": [], | |
393 | "source": [ | |
394 | "e_nn_exo = computeError(data, p_nn_exo, 3)\n", | |
395 | "e_nn_mix = computeError(data, p_nn_mix, 3)\n", | |
396 | "e_az = computeError(data, p_az, 3)\n", | |
397 | "e_pz = computeError(data, p_pz, 3)\n", | |
398 | "options(repr.plot.width=9, repr.plot.height=7)\n", | |
399 | "plotError(list(e_nn_mix, e_pz, e_az, e_nn_exo), cols=c(1,2,colors()[258], 4))\n", | |
400 | "\n", | |
401 | "# Noir: neighbors_mix, bleu: neighbors_exo, vert: moyenne, rouge: persistence\n", | |
402 | "\n", | |
403 | "i_np = which.min(e_nn_exo$abs$indices)\n", | |
404 | "i_p = which.max(e_nn_exo$abs$indices)" | |
405 | ] | |
406 | }, | |
407 | { | |
408 | "cell_type": "code", | |
409 | "execution_count": null, | |
410 | "metadata": { | |
411 | "collapsed": false | |
412 | }, | |
413 | "outputs": [], | |
414 | "source": [ | |
415 | "options(repr.plot.width=9, repr.plot.height=4)\n", | |
416 | "par(mfrow=c(1,2))\n", | |
417 | "\n", | |
418 | "plotPredReal(data, p_nn_exo, i_np); title(paste(\"PredReal nn exo day\",i_np))\n", | |
419 | "plotPredReal(data, p_nn_exo, i_p); title(paste(\"PredReal nn exo day\",i_p))\n", | |
420 | "\n", | |
421 | "plotPredReal(data, p_nn_mix, i_np); title(paste(\"PredReal nn mix day\",i_np))\n", | |
422 | "plotPredReal(data, p_nn_mix, i_p); title(paste(\"PredReal nn mix day\",i_p))\n", | |
423 | "\n", | |
424 | "plotPredReal(data, p_az, i_np); title(paste(\"PredReal az day\",i_np))\n", | |
425 | "plotPredReal(data, p_az, i_p); title(paste(\"PredReal az day\",i_p))\n", | |
426 | "\n", | |
427 | "# Bleu: prévue, noir: réalisée" | |
428 | ] | |
429 | }, | |
430 | { | |
431 | "cell_type": "code", | |
432 | "execution_count": null, | |
433 | "metadata": { | |
434 | "collapsed": false | |
435 | }, | |
436 | "outputs": [], | |
437 | "source": [ | |
438 | "par(mfrow=c(1,2))\n", | |
439 | "f_np_exo = computeFilaments(data, p_nn_exo, i_np, plot=TRUE); title(paste(\"Filaments nn exo day\",i_np))\n", | |
440 | "f_p_exo = computeFilaments(data, p_nn_exo, i_p, plot=TRUE); title(paste(\"Filaments nn exo day\",i_p))\n", | |
441 | "\n", | |
442 | "f_np_mix = computeFilaments(data, p_nn_mix, i_np, plot=TRUE); title(paste(\"Filaments nn mix day\",i_np))\n", | |
443 | "f_p_mix = computeFilaments(data, p_nn_mix, i_p, plot=TRUE); title(paste(\"Filaments nn mix day\",i_p))" | |
444 | ] | |
445 | }, | |
446 | { | |
447 | "cell_type": "code", | |
448 | "execution_count": null, | |
449 | "metadata": { | |
450 | "collapsed": false | |
451 | }, | |
452 | "outputs": [], | |
453 | "source": [ | |
454 | "par(mfrow=c(1,2))\n", | |
455 | "plotFilamentsBox(data, f_np_exo); title(paste(\"FilBox nn exo day\",i_np))\n", | |
456 | "plotFilamentsBox(data, f_p_exo); title(paste(\"FilBox nn exo day\",i_p))\n", | |
457 | "\n", | |
458 | "plotFilamentsBox(data, f_np_mix); title(paste(\"FilBox nn mix day\",i_np))\n", | |
459 | "plotFilamentsBox(data, f_p_mix); title(paste(\"FilBox nn mix day\",i_p))" | |
460 | ] | |
461 | }, | |
462 | { | |
463 | "cell_type": "code", | |
464 | "execution_count": null, | |
465 | "metadata": { | |
466 | "collapsed": false | |
467 | }, | |
468 | "outputs": [], | |
469 | "source": [ | |
470 | "par(mfrow=c(1,2))\n", | |
471 | "plotRelVar(data, f_np_exo); title(paste(\"StdDev nn exo day\",i_np))\n", | |
472 | "plotRelVar(data, f_p_exo); title(paste(\"StdDev nn exo day\",i_p))\n", | |
473 | "\n", | |
474 | "plotRelVar(data, f_np_mix); title(paste(\"StdDev nn mix day\",i_np))\n", | |
475 | "plotRelVar(data, f_p_mix); title(paste(\"StdDev nn mix day\",i_p))\n", | |
476 | "\n", | |
477 | "# Variabilité globale en rouge ; sur les 60 voisins (+ lendemains) en noir" | |
478 | ] | |
479 | }, | |
480 | { | |
481 | "cell_type": "code", | |
482 | "execution_count": null, | |
483 | "metadata": { | |
484 | "collapsed": false | |
485 | }, | |
486 | "outputs": [], | |
487 | "source": [ | |
488 | "par(mfrow=c(1,2))\n", | |
489 | "plotSimils(p_nn_exo, i_np); title(paste(\"Weights nn exo day\",i_np))\n", | |
490 | "plotSimils(p_nn_exo, i_p); title(paste(\"Weights nn exo day\",i_p))\n", | |
491 | "\n", | |
492 | "plotSimils(p_nn_mix, i_np); title(paste(\"Weights nn mix day\",i_np))\n", | |
493 | "plotSimils(p_nn_mix, i_p); title(paste(\"Weights nn mix day\",i_p))\n", | |
494 | "\n", | |
495 | "# - pollué à gauche, + pollué à droite" | |
496 | ] | |
497 | }, | |
498 | { | |
499 | "cell_type": "code", | |
500 | "execution_count": null, | |
501 | "metadata": { | |
502 | "collapsed": false | |
503 | }, | |
504 | "outputs": [], | |
505 | "source": [ | |
506 | "# Fenêtres sélectionnées dans ]0,10] / endo à gauche, exo à droite\n", | |
507 | "p_nn_exo$getParams(i_np)$window\n", | |
508 | "p_nn_exo$getParams(i_p)$window\n", | |
509 | "\n", | |
510 | "p_nn_mix$getParams(i_np)$window\n", | |
511 | "p_nn_mix$getParams(i_p)$window" | |
512 | ] | |
513 | }, | |
514 | { | |
515 | "cell_type": "markdown", | |
516 | "metadata": {}, | |
517 | "source": [ | |
518 | "\n", | |
519 | "\n", | |
520 | "<h2>Bilan</h2>\n", | |
521 | "\n", | |
522 | "Problème difficile : on ne fait guère mieux qu'une naïve moyenne des lendemains des jours\n", | |
523 | "similaires dans le passé, ce qui n'est pas loin de prédire une série constante égale à la\n", | |
524 | "dernière valeur observée (méthode \"zéro\"). La persistence donne parfois de bons résultats\n", | |
525 | "mais est trop instable (sensibilité à l'argument <code>same_day</code>).\n", | |
526 | "\n", | |
527 | "Comment améliorer la méthode ?" | |
528 | ] | |
529 | } | |
530 | ], | |
531 | "metadata": { | |
532 | "kernelspec": { | |
533 | "display_name": "R", | |
534 | "language": "R", | |
535 | "name": "ir" | |
536 | }, | |
537 | "language_info": { | |
538 | "codemirror_mode": "r", | |
539 | "file_extension": ".r", | |
540 | "mimetype": "text/x-r-source", | |
541 | "name": "R", | |
542 | "pygments_lexer": "r", | |
543 | "version": "3.3.3" | |
544 | } | |
545 | }, | |
546 | "nbformat": 4, | |
547 | "nbformat_minor": 2 | |
548 | } |