-
Notifications
You must be signed in to change notification settings - Fork 10
/
Copy pathSuumo.Rmd
542 lines (434 loc) · 18.4 KB
/
Suumo.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
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
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
---
title: "Suumo"
author: "Sho Kosaka"
date: "9/25/2017"
output: html_document
---
```{r}
#必要なライブラリをインポート
library(ggplot2)
library(dplyr)
library(maptools)
library(randomForest)
library(ggrepel)
library(psych)
setwd("~/Desktop/DataAnalysis/Suumo")
df <- read.csv("suumo_tokyo3.csv", fileEncoding = "UTF-16", sep = "\t") #前処理したデータを読み込み
names(df) <- c("index", "name","address","ward","city","rooms","DK","K","L","S","age","height","order","area","rent_adm",
"initial_cost","route1","station1","distance1","route2","station2","distance2","route3","station3","distance3",
"rent","administration","deposit","gratuity","security","non_refund","depreciation") #英語でカラム名付け替え
#factorに変換
df$DK <- as.factor(df$DK)
df$K <- as.factor(df$K)
df$L <- as.factor(df$L)
df$S <- as.factor(df$S)
#df$age_range <- cut(df$age , breaks=c(0,5,10,15,20,25,30,35,40,45,50))
#べき乗で表示されないように
options(scipen=100)
```
```{r}
df1 <- df %>%
select(ward) %>%
group_by(ward) %>%
tally() %>%
arrange((n))
df1$ward <- as.character(df1$ward)
df1$ward <- factor(df1$ward, levels=unique(df1$ward))
df1$pop <- c(58.6,211.3,191.7,143.0,210.3,261.7,219.9,244.0,271.5,280.6,
341.3,452.8,501.5,378.1,321.7,334.2,686.4,678.6,712.1,550.8,
719.1,553.3,883.3)
ggplot(df1,aes(ward,n)) +
geom_bar(stat="identity")+
ggtitle("区別物件数")+xlab("")+ylab("物件数")+
theme(axis.text.x = element_text(angle = 180, hjust =1))+
theme_bw(base_family = "HiraKakuProN-W3") +
coord_flip() +
ylim(0,25000) + scale_y_continuous(labels = scales::comma)
```
```{r}
ggplot(df1,aes(pop,n)) +
geom_point()+
ggtitle("物件数 vs 人口")+xlab("人口（千人）")+ylab("物件数")+
geom_text_repel(label = df1$ward, family = "HiraKakuProN-W3", size = 3) +
theme_bw(base_family = "HiraKakuProN-W3") +
ylim(0,25000) + scale_y_continuous(labels = scales::comma)
```
```{r}
cor(df1$n, df1$pop)
```
```{r}
byrentmedian <- with(df, reorder(ward, rent_adm, median))
par(las=1, cex.axis=0.7, family = "HiraKakuProN-W3")
boxplot(rent_adm ~ byrentmedian, data = df,
xlab = "賃料+管理費", ylab = "",
main = "区別家賃", varwidth = TRUE, horizontal=TRUE)
```
```{r}
par(las=1, cex.axis=0.7, family = "HiraKakuProN-W3")
boxplot(rent_adm ~ byrentmedian, data = df,
xlab = "賃料+管理費", ylab = "",
main = "区別家賃（〜40万円）", varwidth = TRUE, horizontal=TRUE, ylim = c(0,400000), outline=FALSE)
```
```{r}
ggplot(df,aes(rooms, rent_adm)) + geom_jitter() +ggtitle("賃料+管理費 vs 部屋数")+xlab("部屋数")+ylab("賃料+管理費")+ theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3") + geom_smooth(method="lm") + scale_x_continuous(breaks=seq(0,10,1)) + scale_y_continuous(labels = scales::comma)
```
```{r}
#エラーデータ削除
df <- subset(df, rooms <=9)
subset(df, rooms >= 9)
```
```{r}
#不要
ggplot(df,aes(rent_adm)) + geom_freqpoly(bins = 100) +ggtitle("賃料+管理費 分布")+xlab("賃料+管理費")+ylab("物件数")+ theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3") + xlim(0,500000)
```
```{r}
df_sub <- subset(df, ward=="港区" | ward=="世田谷区" | ward=="葛飾区")
ggplot(df_sub,aes(rent_adm, color = ward)) + geom_freqpoly(bins = 100) +ggtitle("賃料+管理費 分布（世田谷区、港区、葛飾区）")+xlab("賃料+管理費")+ylab("物件数")+ theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3") + scale_x_continuous(labels = scales::comma, limits=c(0, 500000)) + scale_y_continuous(labels = scales::comma)
```
```{r}
ggplot(df) +
geom_freqpoly(aes(distance1, color="blue"),bins = 30)+
geom_freqpoly(aes(distance2, color="green"),bins = 30)+
geom_freqpoly(aes(distance3, color="red"),bins = 30)+
ggtitle("最寄り３駅からの距離")+xlab("徒歩〜分")+ylab("物件数")+
theme_bw(base_family = "HiraKakuProN-W3")+
xlim(0,30) +scale_y_continuous(labels = scales::comma) +
scale_color_hue(name = "", labels = c("一つ目の立地", "二つ目の立地","三つ目の立地"))
```
```{r}
ggplot(df,aes(x=rent_adm,y=initial_cost)) +
geom_point()+
ggtitle("敷金・礼金・保証金 vs 賃料")+xlab("賃料")+ylab("敷金・礼金・保証金")+
theme_bw(base_family = "HiraKakuProN-W3") +scale_y_continuous(labels = scales::comma, limits = c(0,1000000))+
scale_x_continuous(labels = scales::comma, limits = c(0,500000))
```
```{r}
byagemedian <- with(df, reorder(ward, age, median))
par(las=1, cex.axis=0.7, family = "HiraKakuProN-W3")
boxplot(age ~ byagemedian, data = df,
xlab = "築年数", ylab = "",
main = "区別築年数", varwidth = TRUE, horizontal=TRUE, ylim = c(0,70), outline=FALSE)
```
```{r}
#不要
ggplot(df,aes(x=area,y=rooms)) +
geom_jitter()+
ggtitle("面積 vs 部屋数")+xlab("面積")+ylab("部屋数")+
theme_bw(base_family = "HiraKakuProN-W3")+scale_y_continuous(breaks=seq(0,10,1)) + xlim(0,500)
```
```{r}
df11 <- df %>%
select(rent_adm, ward, rooms,area, DK, K, L, S, age, height, order, distance1)
pairs.panels(df11)
```
```{r}
df12 <- df %>%
select(rent_adm, ward, rooms, K, L, S, age, height, distance1)
pairs.panels(df12)
```
```{r}
set.seed(1234)
id_train <- sample(1:nrow(df),nrow(df)/5*4)
train <- df[id_train,]
test <- df[-id_train,]
```
```{r}
low <- 100000
high <- 200000
df_low <- subset(df, rent < low)
df_middle <- subset(df, rent >= low & rent < high)
df_high <- subset(df, rent >= high)
```
```{r}
set.seed(1234)
id_train1 <- sample(1:nrow(df_low),nrow(df_low)/5*4)
train1 <- df_low[id_train1,]
test1 <- df_low[-id_train1,]
lm_rent_low <- lm(rent_adm ~ ward + rooms + K + L + S + age + height + distance1
#+distance2 + distance3
, data = train1)
#summary(lm_rent_low)
plot(lm_rent_low, which=2)
```
```{r}
set.seed(1234)
id_train2 <- sample(1:nrow(df_middle),nrow(df_middle)/5*4)
train2 <- df_middle[id_train2,]
test2 <- df_middle[-id_train2,]
lm_rent_middle <- lm(rent_adm ~ ward + rooms + K + L + S + age + height + distance1
#+distance2 + distance3
, data = train2)
#summary(lm_rent_middle)
plot(lm_rent_middle, which=2)
```
```{r}
set.seed(1234)
id_train3 <- sample(1:nrow(df_high),nrow(df_high)/5*4)
train3 <- df_high[id_train3,]
test3 <- df_high[-id_train3,]
lm_rent_high <- lm(rent_adm ~ ward + rooms + K + L + S + age + height + distance1
#+distance2 + distance3
, data = train3)
#summary(lm_rent_high)
plot(lm_rent_high, which=2)
```
```{r}
#areaはroomsと、DKはKと、orderはheightと相関が高いため除外
lm_rent <- lm(rent_adm ~ ward + rooms + K + L + S + age + height + distance1
#+distance2 + distance3
, data = train)
summary(lm_rent)
sink()
barplot(lm_rent$coefficients)
plot(lm_rent, which=2)
```
```{r}
lm_rent1 <- lm(rent_adm ~ (ward + rooms + DK + K + L + S + area + age + height + order + distance1)^2
#+distance2 + distance3
, data = train)
lm_rent2 <- step(lm_rent1)
saveRDS(lm_rent2, file="lm_rent2")
lm_rent2 <- readRDS(file="lm_rent2")
lm_rent2
```
```{r}
df_lm_pred <- predict(lm_rent, newdata = test)
```
```{r}
df7 <- df %>%
group_by(ward) %>%
summarise(rent_mean = mean(rent_adm))
for (x in 1:nrow(df7)) {
if (df7$ward[x] == "足立区") {
df7$ward_eng[x] <- "Adachi-ku"
} else if (df7$ward[x] == "荒川区") {
df7$ward_eng[x] <- "Arakawa-ku"
} else if (df7$ward[x] == "文京区") {
df7$ward_eng[x] <- "Bunkyo-ku"
} else if (df7$ward[x] == "千代田区") {
df7$ward_eng[x] <- "Chiyoda-ku"
} else if (df7$ward[x] == "中央区") {
df7$ward_eng[x] <- "Chuo-ku"
} else if (df7$ward[x] == "江戸川区") {
df7$ward_eng[x] <- "Edogawa-ku"
} else if (df7$ward[x] == "板橋区") {
df7$ward_eng[x] <- "Itabashi-ku"
} else if (df7$ward[x] == "葛飾区") {
df7$ward_eng[x] <- "Katsushika-ku"
} else if (df7$ward[x] == "北区") {
df7$ward_eng[x] <- "Kita-ku"
} else if (df7$ward[x] == "江東区") {
df7$ward_eng[x] <- "Koto-ku"
} else if (df7$ward[x] == "目黒区") {
df7$ward_eng[x] <- "Meguro-ku"
} else if (df7$ward[x] == "港区") {
df7$ward_eng[x] <- "Minato-ku"
} else if (df7$ward[x] == "中野区") {
df7$ward_eng[x] <- "Nakano-ku"
} else if (df7$ward[x] == "練馬区") {
df7$ward_eng[x] <- "Nerima-ku"
} else if (df7$ward[x] == "大田区") {
df7$ward_eng[x] <- "Ota-ku"
} else if (df7$ward[x] == "世田谷区") {
df7$ward_eng[x] <- "Setagaya-ku"
} else if (df7$ward[x] == "渋谷区") {
df7$ward_eng[x] <- "Shibuya-ku"
} else if (df7$ward[x] == "品川区") {
df7$ward_eng[x] <- "Shinagawa-ku"
} else if (df7$ward[x] == "新宿区") {
df7$ward_eng[x] <- "Shinjuku-ku"
} else if (df7$ward[x] == "杉並区") {
df7$ward_eng[x] <- "Suginami-ku"
} else if (df7$ward[x] == "墨田区") {
df7$ward_eng[x] <- "Sumida-ku"
} else if (df7$ward[x] == "台東区") {
df7$ward_eng[x] <- "Taito-ku"
} else if (df7$ward[x] == "豊島区") {
df7$ward_eng[x] <- "Toshima-ku"
}
}
```
```{r}
#http://ill-identified.hatenablog.com/entry/2014/02/27/211021を参考にした
tokyo.map <- readShapeSpatial("~/Desktop/Python/Suumo/japan_ver81.shp")
# パスは各自任意で変更すること.
# シェイプファイルに含まれる属性データは以下の通り.
# N03_001: 都道府県名
# N03_002: 支庁・振興局名 (北海道のみ)
# N03_003: 郡・政令市名
# N03_004: 市区町村名
# N03_007: 行政区域コード
# fortify で data.frame に変換した後の結合キーにするため, rowname を変数に代入
# 参考: http://www.r-bloggers.com/using-r-working-with-geospatial-data-and-ggplot2/
plot(tokyo.map)
tokyo.map$id <- rownames(tokyo.map@data)
tokyo.map <- subset(tokyo.map, CITY_ENG == "Adachi-ku" | CITY_ENG == "Arakawa-ku" | CITY_ENG == "Bunkyo-ku" |
CITY_ENG == "Chiyoda-ku" | CITY_ENG == "Chuo-ku" | CITY_ENG == "Edogawa-ku" |
CITY_ENG == "Itabashi-ku" | CITY_ENG == "Katsushika-ku" | CITY_ENG == "Kita-ku" |
CITY_ENG == "Koto-ku" | CITY_ENG == "Meguro-ku" | CITY_ENG == "Minato-ku" |
CITY_ENG == "Nakano-ku" | CITY_ENG == "Nerima-ku" | CITY_ENG == "Ota-ku" |
CITY_ENG == "Setagaya-ku" | CITY_ENG == "Shibuya-ku" | CITY_ENG == "Shinagawa-ku" |
CITY_ENG == "Shinjuku-ku" | CITY_ENG == "Suginami-ku" | CITY_ENG == "Sumida-ku" |
CITY_ENG == "Taito-ku" | CITY_ENG == "Toshima-ku" )
# 属性データに選挙情報を結合
# 市区町村名をキーとして左結合
tokyo.map2 <- merge(x=tokyo.map, y=df7, by.x="CITY_ENG",by.y="ward_eng", duplicateGeoms=TRUE)
# ポリゴンデータをデータフレームに変換
tokyo.gg.df <- fortify(tokyo.map2)
# データフレーム化したポリゴンデータに属性データを左結合
tokyo.gg.df <- merge(x=tokyo.gg.df, y=tokyo.map2@data, by="id", all.x = T)
# 区名ラベル位置座標
location <- lapply(tokyo.map2@polygons, function(x) slot(x,"labpt") )
location <- as.data.frame(matrix(unlist(location), ncol=2, byrow=T, dimnames=list(NULL,c("x","y"))) )
location$id <-unlist(lapply(tokyo.map2@polygons, function(x) slot(x,"ID") ) )
# ラベル位置と市区町村名結合
label.gg.df <- merge(x=location, y=unique(tokyo.gg.df[,c("CITY_ENG","id")]), by.x="id", by.y="id", all.x = T)
label.gg.df <- label.gg.df[!duplicated(label.gg.df$CITY_ENG),] # 重複削除 (雑)
#write.csv(tokyo.gg.df, file = "tokyo.gg.df.csv")
#tokyo.gg.df1 <- read.csv("tokyo.gg.df.csv", fileEncoding = "UTF-8", sep = "\t")
write.csv(label.gg.df, file = "label.gg.df.csv")
label.gg.df1 <- read.csv("label.gg.df.csv", fileEncoding = "UTF-8", sep = ",") #Excel上で手動で日本語化。UTF-8で保存。
ggplot(tokyo.gg.df) +
geom_polygon(aes(long,lat, group=group, fill=rent_mean), color="white") +
geom_text(data =label.gg.df1, aes(x,y,label=CITY_ENG, family = "HiraKakuPro-W3"), size=3) +
scale_fill_continuous(name="平均賃料") +
labs(title="東京23区 平均賃料")+theme_bw(base_family = "HiraKakuProN-W3")
```
```{r}
#不使用
library(ggmap)
map <- get_map(c(139.6917, 35.68949))
map <- get_googlemap(center=c(139.6917,35.68949), scale=2, size=c(600, 600), zoom=12)
ggmap(map)
ggmap(map) + geom_point(data=c(139.7238,35.64918), aes(x=経度, y=緯度), color='red')
```
```{r}
#不要
center <- geocode('kanazawa')
map <- get_map(c(center$lon, center$lat), zoom = 5, maptype = 'roadmap')
ggmap(map) + geom_point(data=locations, aes(x = geo_coordinates_1, y = geo_coordinates_0), color = 'red', size = 0.7)
```
```{r}
#16時間かかるので注意
df_rf <- randomForest(rent_adm ~ ward + rooms + area + DK + K + L + S + age + height + order + distance1
#+distance2 + distance3
, data = train, na.action = "na.omit", trees = 300)
saveRDS(df_rf, file="df_rf")
df_rf <- readRDS(file="df_rf")
print(df_rf)
summary(df_rf)
plot(df_rf)
print(df_rf$importance)
label_rf <- c("区","部屋数","L","建物高さ","階数","築年数","DK","徒歩〜分","S","K")
varImpPlot(df_rf, main="Contribution")
```
```{r}
features <- train[c(4,6:14)]
labels <- train[15]
tuneRF(x =features, y = train$rent_adm,
#stepFactor = 2, improve = 0.05, trace = TRUE, plot = TRUE,
doBest = TRUE )
str(train)
```
```{r}
df_rf_pred <- predict(df_rf, test[,-15]) #rent_admを削除
df_rf_pred_csv <- data.frame(df_rf_pred)
write.csv(df_rf_pred_csv, "df_rf_pred.csv")
```
```{r}
ggplot(test) + geom_line(aes(x=1:nrow(test),y=rent_adm,color="red"))+
#geom_line(aes(x=1:nrow(test),y=df_rf_pred,color="red"))+
#geom_line(aes(x=1:nrow(test),y=df_lm_pred,color="blue"))+
ggtitle("テストデータ")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
ggplot(test) + geom_line(aes(x=1:nrow(test),y=rent_adm,color="blue"))+
#geom_line(aes(x=1:nrow(test),y=df_rf_pred,color="red"))+
geom_line(aes(x=1:nrow(test),y=df_lm_pred,color="red"))+
ggtitle("テストデータ vs 予測データ（重回帰分析）")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ","重回帰分析") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
ggplot(test) + geom_line(aes(x=1:nrow(test),y=rent_adm,color="blue"))+
geom_line(aes(x=1:nrow(test),y=df_rf_pred,color="red"))+
#geom_line(aes(x=1:nrow(test),y=df_lm_pred,color="green"))+
ggtitle("テストデータ vs 予測データ（ランダムフォレスト）")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ","ランダムフォレスト") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
test_sort <- test %>%
arrange((rent_adm))
ggplot(test_sort) + geom_line(aes(x=1:nrow(test_sort),y=rent_adm,color="red"))+
#geom_line(aes(x=1:nrow(test),y=df_rf_pred,color="red"))+
#geom_line(aes(x=1:nrow(test),y=df_lm_pred,color="blue"))+
ggtitle("テストデータ")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
df_lm_pred_sort <- predict(lm_rent, newdata = test_sort)
ggplot(test_sort) +
#geom_line(aes(x=1:nrow(test),y=df_rf_pred,color="red"))+
geom_line(aes(x=1:nrow(test_sort),y=df_lm_pred_sort,color="red"))+
geom_line(aes(x=1:nrow(test_sort),y=rent_adm,color="blue"))+
ggtitle("テストデータ vs 予測データ（重回帰分析）")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ","重回帰分析") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
df_rf_pred_sort <- predict(df_rf, test_sort[,-15])
ggplot(test_sort) +
geom_line(aes(x=1:nrow(test_sort),y=df_rf_pred_sort,color="red"))+
geom_line(aes(x=1:nrow(test_sort),y=rent_adm,color="blue"))+
#geom_line(aes(x=1:nrow(test),y=df_lm_pred,color="green"))+
ggtitle("テストデータ vs 予測データ（ランダムフォレスト）")+xlab("index")+ylab("家賃")+
scale_color_hue(name = "", labels = c("テストデータ","ランダムフォレスト") ) +
theme(axis.text.x = element_text(angle = 180, hjust =1))+theme_bw(base_family = "HiraKakuProN-W3")+
ylim(0,1000000)
```
```{r}
df_rf_diff <- df_rf_pred - test$rent_adm
df_lm_diff <- df_lm_pred - test$rent_adm
df_rf_diff_sq <- df_rf_diff ^ 2
df_lm_diff_sq <- df_lm_diff ^ 2
sum(df_rf_diff_sq, na.rm = TRUE)
sum(df_lm_diff_sq, na.rm = TRUE)
table_diff <- data.frame("diff_sq" = c(sum(df_rf_diff_sq, na.rm = TRUE),sum(df_lm_diff_sq, na.rm = TRUE)),
"model" = c("ランダムフォレスト","重回帰分析"))
ggplot(table_diff,aes(model,diff_sq)) +
geom_bar(stat = "identity")+
ggtitle("残差平方和：ランダムフォレスト vs 重回帰分析")+xlab("")+ylab("残差平方和")+
theme_bw(base_family = "HiraKakuProN-W3") + scale_y_continuous(labels = scales::comma)
```
```{r}
df$pred_rf <- predict(df_rf, newdata = df)
df$diff <- df$pred_rf - df$rent_adm
df_sort <- df %>%
arrange(desc(diff))
```
```{r}
df_tokyo2 <- read.csv("suumo_tokyo2.csv", fileEncoding = "UTF-16", sep = "\t") #前処理したデータを読み込み
names(df_tokyo2) <- c("index", "name","address","ward","city","rooms","DK","K","L","S","age","height","order","area","rent_adm",
"initial_cost","route1","station1","distance1","route2","station2","distance2","route3","station3","distance3",
"rent","administration","deposit","gratuity","security","non_refund","depreciation") #英語でカラム名付け替え
#factorに変換
df_tokyo2$DK <- as.factor(df_tokyo2$DK)
df_tokyo2$K <- as.factor(df_tokyo2$K)
df_tokyo2$L <- as.factor(df_tokyo2$L)
df_tokyo2$S <- as.factor(df_tokyo2$S)
df_tokyo2$pred_rf <- predict(df_rf, newdata = df_tokyo2)
df_tokyo2$diff <- df_tokyo2$pred_rf - df_tokyo2$rent_adm
df_tokyo2_sort <- df_tokyo2 %>%
arrange(desc(diff))
df_tokyo2_50 <- df_tokyo2_sort[1:50,]
readr::write_excel_csv(df_tokyo2_50, "df_tokyo2_50.csv")
```