From 97c99c8e9ded13ee93878356a435ad7ea8c6017d Mon Sep 17 00:00:00 2001
From: Petr Baudis <pasky@ucw.cz>
Date: Sat, 13 Mar 2010 20:44:28 +0100
Subject: [PATCH] tex: Plot of number of pattern occurences

---
 tex/gostyle.tex       |  12 +-
 tex/makefile          |   5 +-
 tex/patcountdist.data | 500 ++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 515 insertions(+), 2 deletions(-)
 create mode 100644 tex/patcountdist.data

diff --git a/tex/gostyle.tex b/tex/gostyle.tex
index c7cf07f..0631fa7 100644
--- a/tex/gostyle.tex
+++ b/tex/gostyle.tex
@@ -436,6 +436,14 @@ Thus, we obtain vectors describing relative frequency of played patterns
 independent on number of gathered patterns.
 But there are multiple ways to approach the normalization.
 
+\begin{figure}[!t]
+\centering
+\includegraphics{patcountdist}
+\caption{Log-scaled number of pattern occurences
+in the GoGoD games examined in sec. \ref{styleest}.}
+\label{fig:patcountdist}
+\end{figure}
+
 \subsubsection{Linear Normalization}
 
 One is simply to linearly re-scale the values using:
@@ -460,7 +468,9 @@ However, we have found that this method is not universally beneficial.
 In our styles case study (sec. \ref{styleest}), this normalization
 produced PCA decomposition with significant dimensions corresponding
 better to some of the prior knowledge and more instructive for manual
-inspection, but ultimately worsened accuracy of our classifiers.
+inspection, but ultimately worsened accuracy of our classifiers;
+we conjecture from this that the most frequently occuring patterns are
+also most important for classification of major style aspects.
 
 \subsection{Implementation}
 
diff --git a/tex/makefile b/tex/makefile
index b104470..2bc9dee 100644
--- a/tex/makefile
+++ b/tex/makefile
@@ -1,4 +1,4 @@
-gostyle.dvi: gostyle.tex gostyle.bib strength-pca.eps style-pca.eps makefile
+gostyle.dvi: gostyle.tex gostyle.bib patcountdist.eps strength-pca.eps style-pca.eps makefile
 	rm -f gostyle.bbl
 	latex gostyle && bibtex gostyle && latex gostyle && latex gostyle
 
@@ -8,6 +8,9 @@ gostyle.pdf: gostyle.dvi
 gostyle.ps: gostyle.dvi
 	dvips gostyle.dvi
 
+patcountdist.eps: patcountdist.data
+	echo 'set term postscript eps enhanced "Helvetica" 18 size 3.5,2.5; set lmargin 7; set xlabel "pattern id"; set output "'$@'"; set log y; set xtics 1 125; plot "'$<'" notitle pointtype 2' | gnuplot
+
 strength-pca.eps: strength-pca.data
 	echo 'set term postscript eps enhanced "Helvetica" 18 size 7,2; set output "'$@'"; set xtics -4 2; set xlabel "rank (-d, +k)" ; set ylabel "first dim." ; plot "'$<'" using 2:3 notitle pointtype 2' | gnuplot
 
diff --git a/tex/patcountdist.data b/tex/patcountdist.data
new file mode 100644
index 0000000..d2f752f
--- /dev/null
+++ b/tex/patcountdist.data
@@ -0,0 +1,500 @@
+18971
+18033
+15753
+8005
+5033
+4435
+4097
+2924
+2598
+2385
+1906
+1905
+1591
+1580
+1579
+1517
+1418
+1413
+1390
+1364
+1258
+1235
+1139
+1133
+1108
+1084
+1028
+991
+980
+977
+948
+943
+900
+869
+858
+853
+837
+821
+810
+778
+773
+748
+743
+738
+722
+678
+666
+650
+647
+645
+636
+631
+610
+608
+606
+599
+598
+598
+592
+591
+588
+572
+565
+565
+554
+553
+539
+528
+523
+523
+513
+513
+511
+509
+508
+507
+506
+504
+502
+499
+499
+487
+485
+484
+475
+473
+472
+468
+468
+467
+465
+462
+458
+457
+454
+448
+448
+443
+441
+438
+437
+434
+423
+422
+422
+420
+420
+414
+414
+410
+410
+408
+407
+406
+404
+399
+397
+396
+395
+394
+393
+392
+391
+391
+391
+389
+388
+387
+383
+383
+382
+379
+376
+373
+373
+372
+371
+370
+368
+368
+367
+366
+363
+362
+361
+360
+360
+358
+357
+356
+355
+354
+354
+354
+353
+353
+350
+350
+348
+347
+346
+345
+342
+341
+339
+337
+336
+336
+336
+333
+331
+331
+329
+329
+326
+324
+321
+318
+317
+317
+316
+314
+313
+313
+311
+310
+309
+309
+307
+306
+305
+305
+305
+304
+303
+301
+301
+300
+299
+298
+298
+297
+297
+296
+295
+294
+293
+293
+292
+292
+292
+291
+291
+291
+290
+289
+289
+288
+285
+285
+285
+285
+285
+283
+283
+283
+283
+283
+282
+282
+281
+280
+280
+279
+279
+279
+278
+276
+276
+276
+275
+275
+275
+274
+274
+273
+273
+273
+273
+273
+272
+272
+271
+271
+271
+270
+269
+269
+268
+268
+268
+268
+267
+265
+265
+264
+264
+263
+263
+263
+262
+262
+262
+262
+261
+261
+261
+261
+259
+259
+259
+259
+258
+258
+258
+257
+257
+257
+257
+256
+255
+255
+255
+255
+254
+254
+254
+254
+254
+254
+254
+254
+253
+253
+253
+252
+252
+252
+252
+251
+250
+250
+250
+250
+249
+249
+249
+248
+248
+248
+248
+247
+247
+247
+247
+247
+246
+246
+246
+246
+246
+246
+245
+245
+245
+245
+244
+243
+243
+243
+243
+242
+242
+242
+241
+241
+241
+241
+240
+240
+240
+240
+239
+239
+239
+238
+238
+238
+237
+237
+236
+236
+235
+235
+234
+233
+233
+233
+233
+231
+230
+230
+230
+230
+229
+229
+228
+228
+228
+228
+228
+228
+228
+228
+227
+227
+227
+227
+227
+227
+226
+226
+226
+226
+225
+225
+225
+224
+224
+224
+223
+223
+223
+223
+223
+223
+222
+222
+222
+222
+222
+221
+221
+221
+221
+221
+221
+220
+220
+220
+220
+220
+219
+219
+219
+219
+219
+218
+218
+218
+218
+218
+217
+216
+216
+215
+215
+215
+215
+215
+215
+214
+214
+214
+214
+214
+214
+213
+213
+213
+213
+213
+212
+212
+212
+212
+211
+211
+210
+210
+210
+210
+210
+210
+209
+209
+209
+209
+209
+208
+208
+208
+208
+208
+208
+208
+208
+207
+207
+207
+207
+207
+206
+206
+206
+205
+205
+204
+204
+204
+204
+204
+204
+204
+203
+203
+203
+203
+203
+203
-- 
2.11.4.GIT