安井 翔太（著）『効果検証入門』（技術評論社）を読んでいて、Rのパイプ関数（%>%）が使い勝手が良さそうだったので、写経して使ってみた。
###（ 関連記事 ）
#実行環境
- RStudio
- OS: macOS Catalina
なお、別の記事では、この記事と同じコードを、Google Colaboratory上でも実行してみました。
詳細は次の記事を参照ください。
##パッケージのインストールと読込み
RStudio
> install.packages("tidyverse")
（ ・・・省略・・・ ）
>
RStudio
> library("tidyverse")
─ Attaching packages ───────────────────── tidyverse 1.3.0 ─
✓ ggplot2 3.3.3 ✓ purrr 0.3.4
✓ tibble 3.0.4 ✓ dplyr 1.0.2
✓ tidyr 1.1.2 ✓ stringr 1.4.0
✓ readr 1.4.0 ✓ forcats 0.5.0
─ Conflicts ────────────────────── tidyverse_conflicts() ─
x dplyr::filter() masks stats::filter()
x dplyr::lag() masks stats::lag()
>
##データセットのダウンロードと読込み
RStudio
> csv_url = "http://www.minethatdata.com/Kevin_Hillstrom_MineThatData_E-MailAnalytics_DataMiningChallenge_2008.03.20.csv"
> email_data <- read_csv(csv_url)
─ Column specification ─────────────────────────────
cols(
recency = col_double(),
history_segment = col_character(),
history = col_double(),
mens = col_double(),
womens = col_double(),
zip_code = col_character(),
newbie = col_double(),
channel = col_character(),
segment = col_character(),
visit = col_double(),
conversion = col_double(),
spend = col_double()
)
>
####データのカラム構造と分布の確認
RStudio
> summary(email_data)
recency history_segment history mens
Min. : 1.000 Length:64000 Min. : 29.99 Min. :0.000
1st Qu.: 2.000 Class :character 1st Qu.: 64.66 1st Qu.:0.000
Median : 6.000 Mode :character Median : 158.11 Median :1.000
Mean : 5.764 Mean : 242.09 Mean :0.551
3rd Qu.: 9.000 3rd Qu.: 325.66 3rd Qu.:1.000
Max. :12.000 Max. :3345.93 Max. :1.000
womens zip_code newbie channel
Min. :0.0000 Length:64000 Min. :0.0000 Length:64000
1st Qu.:0.0000 Class :character 1st Qu.:0.0000 Class :character
Median :1.0000 Mode :character Median :1.0000 Mode :character
Mean :0.5497 Mean :0.5022
3rd Qu.:1.0000 3rd Qu.:1.0000
Max. :1.0000 Max. :1.0000
segment visit conversion spend
Length:64000 Min. :0.0000 Min. :0.000000 Min. : 0.000
Class :character 1st Qu.:0.0000 1st Qu.:0.000000 1st Qu.: 0.000
Mode :character Median :0.0000 Median :0.000000 Median : 0.000
Mean :0.1468 Mean :0.009031 Mean : 1.051
3rd Qu.:0.0000 3rd Qu.:0.000000 3rd Qu.: 0.000
Max. :1.0000 Max. :1.000000 Max. :499.000
>
RStudio
> head(email_data)
# A tibble: 6 x 12
recency history_segment history mens womens zip_code newbie channel segment
<dbl> <chr> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <chr>
1 10 2) $100 - $200 142. 1 0 Surburb… 0 Phone Womens…
2 6 3) $200 - $350 329. 1 1 Rural 1 Web No E-M…
3 7 2) $100 - $200 181. 0 1 Surburb… 1 Web Womens…
4 9 5) $500 - $750 676. 1 0 Rural 1 Web Mens E…
5 2 1) $0 - $100 45.3 1 0 Urban 0 Web Womens…
6 6 2) $100 - $200 135. 0 1 Surburb… 0 Phone Womens…
# … with 3 more variables: visit <dbl>, conversion <dbl>, spend <dbl>
>
RStudio
> tail(email_data)
# A tibble: 6 x 12
recency history_segment history mens womens zip_code newbie channel segment
<dbl> <chr> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <chr>
1 7 1) $0 - $100 86.5 0 1 Urban 0 Web Mens E…
2 10 2) $100 - $200 106. 1 0 Urban 0 Web Mens E…
3 5 1) $0 - $100 38.9 0 1 Urban 1 Phone Mens E…
4 6 1) $0 - $100 30.0 1 0 Urban 1 Phone Mens E…
5 1 5) $500 - $750 553. 1 0 Surburb… 1 Multic… Womens…
6 1 4) $350 - $500 473. 0 1 Surburb… 0 Web Mens E…
# … with 3 more variables: visit <dbl>, conversion <dbl>, spend <dbl>
>
##__email_dataをデータ加工する処理を、%>%__記法でパイプで繋いで、実行
なお、__"+"__は、RStudioでコードの途中で改行した際に付される記号です。
コードに*"+"*を記述する必要はありません。
RStudio
> male_df <- email_data %>%
+ filter(segment != "Womens E-Mail")
>
RStudio
> male_df
# A tibble: 42,613 x 12
recency history_segment history mens womens zip_code newbie channel segment
<dbl> <chr> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <chr>
1 6 3) $200 - $350 329. 1 1 Rural 1 Web No E-M…
2 9 5) $500 - $750 676. 1 0 Rural 1 Web Mens E…
3 9 5) $500 - $750 675. 1 1 Rural 1 Phone Mens E…
4 2 2) $100 - $200 102. 0 1 Urban 0 Web Mens E…
5 4 3) $200 - $350 241. 0 1 Rural 1 Multic… No E-M…
6 3 1) $0 - $100 58.1 1 0 Urban 1 Web No E-M…
7 5 1) $0 - $100 30.0 1 0 Surburb… 0 Phone Mens E…
8 9 2) $100 - $200 112. 1 0 Rural 0 Web Mens E…
9 11 3) $200 - $350 219. 1 1 Surburb… 0 Phone Mens E…
10 5 6) $750 - $1,0… 828. 1 0 Surburb… 1 Multic… Mens E…
# … with 42,603 more rows, and 3 more variables: visit <dbl>, conversion <dbl>,
# spend <dbl>
>
RStudio
> male_df <- email_data %>%
+ filter(segment != "Womens E-Mail") %>%
+ mutate(treatment = if_else(segment == "Mens E-Mail", 1, 0))
>
RStudio
> summary(male_df)
recency history_segment history mens
Min. : 1.000 Length:42613 Min. : 29.99 Min. :0.0000
1st Qu.: 2.000 Class :character 1st Qu.: 64.50 1st Qu.:0.0000
Median : 6.000 Mode :character Median : 157.00 Median :1.0000
Mean : 5.762 Mean : 241.86 Mean :0.5521
3rd Qu.: 9.000 3rd Qu.: 325.21 3rd Qu.:1.0000
Max. :12.000 Max. :3345.93 Max. :1.0000
womens zip_code newbie channel
Min. :0.0000 Length:42613 Min. :0.0000 Length:42613
1st Qu.:0.0000 Class :character 1st Qu.:0.0000 Class :character
Median :1.0000 Mode :character Median :1.0000 Mode :character
Mean :0.5495 Mean :0.5017
3rd Qu.:1.0000 3rd Qu.:1.0000
Max. :1.0000 Max. :1.0000
segment visit conversion spend
Length:42613 Min. :0.0000 Min. :0.000000 Min. : 0.000
Class :character 1st Qu.:0.0000 1st Qu.:0.000000 1st Qu.: 0.000
Mode :character Median :0.0000 Median :0.000000 Median : 0.000
Mean :0.1445 Mean :0.009129 Mean : 1.038
3rd Qu.:0.0000 3rd Qu.:0.000000 3rd Qu.: 0.000
Max. :1.0000 Max. :1.000000 Max. :499.000
treatment
Min. :0.0
1st Qu.:0.0
Median :1.0
Mean :0.5
3rd Qu.:1.0
Max. :1.0
>
> head(male_df)
# A tibble: 6 x 13
recency history_segment history mens womens zip_code newbie channel segment
<dbl> <chr> <dbl> <dbl> <dbl> <chr> <dbl> <chr> <chr>
1 6 3) $200 - $350 329. 1 1 Rural 1 Web No E-M…
2 9 5) $500 - $750 676. 1 0 Rural 1 Web Mens E…
3 9 5) $500 - $750 675. 1 1 Rural 1 Phone Mens E…
4 2 2) $100 - $200 102. 0 1 Urban 0 Web Mens E…
5 4 3) $200 - $350 241. 0 1 Rural 1 Multic… No E-M…
6 3 1) $0 - $100 58.1 1 0 Urban 1 Web No E-M…
# … with 4 more variables: visit <dbl>, conversion <dbl>, spend <dbl>,
# treatment <dbl>
> head(male_df$treatment)
[1] 0 1 1 1 0 0
>
> tail(male_df$treatment)
[1] 1 1 1 1 1 1
>
> male_df$treatment
[1] 0 1 1 1 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 0 1 1 1 1 1 1 1 1 0 0 0 0 1 0 1 1 1 0
[39] 0 0 1 1 1 0 0 0 1 1 1 0 1 0 0 1 1 1 0 1 1 0 0 1 1 1 1 0 1 1 0 0 1 0 1 1 0 1
[77] 0 1 0 0 1 1 1 0 1 0 1 1 0 0 1 0 1 1 1 1 1 1 0 0 1 0 0 0 0 0 1 1 0 1 1 0 1 1
[115] 1 0 0 1 0 0 1 1 1 0 1 1 0 1 0 0 0 0 1 0 0 0 0 1 0 0 1 0 1 1 0 1 0 0 1 0 1 0
[153] 0 1 0 1 0 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 0 1 0
[191] 0 1 0 0 1 1 0 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 0 1 0 1 0 1 0 1 0 1 1 1 1 0
[229] 0 0 1 1 0 1 0 0 1 0 0 0 1 0 0 1 0 1 1 1 1 0 0 1 0 1 1 0 1 0 0 1 0 0 0 0 1 0
[267] 0 1 0 1 0 1 0 1 0 0 0 0 0 1 1 0 1 1 1 1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 0 0 0
[305] 0 1 0 0 0 0 0 1 1 0 0 0 1 1 1 0 1 1 0 1 1 0 1 0 1 0 0 0 1 1 0 0 1 0 0 0 0 0
[343] 0 0 0 1 0 1 0 0 1 1 1 0 1 1 1 1 0 1 0 0 0 1 0 0 0 0 1 1 1 1 0 0 1 0 0 1 1 0
[381] 1 1 1 1 1 0 1 1 1 1 0 0 1 1 1 0 0 1 1 0 0 0 1 0 1 0 1 0 0 0 1 0 0 0 0 0 0 0
[419] 1 0 0 0 1 1 0 1 0 0 0 1 1 1 0 0 1 0 1 1 1 0 0 0 0 1 1 0 0 0 0 0 1 1 1 1 1 0
[457] 0 0 1 1 0 1 1 1 0 0 1 1 0 1 0 0 1 0 0 0 0 0 1 0 0 0 1 1 0 0 1 1 0 1 1 0 1 1
[495] 0 1 0 0 0 0 0 1 1 0 0 1 1 0 0 0 0 1 1 1 0 1 1 0 0 0 1 1 0 0 0 1 1 1 0 1 1 0
[533] 1 0 0 1 0 0 1 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 0 1 0 1 0 0 1 0 0 1 1 1 1 0 1 1
[571] 1 0 1 0 0 0 0 1 1 1 0 0 0 0 1 1 1 0 1 0 0 1 0 1 1 0 1 1 1 0 0 1 1 0 1 0 0 0
[609] 0 1 1 0 0 1 0 1 1 1 1 0 1 1 0 0 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 1 0 1 0 0 1 0
[647] 1 1 0 0 1 1 0 1 0 1 1 0 0 1 1 0 1 1 0 0 0 0 1 0 1 0 0 0 0 1 1 0 0 1 0 0 1 0
[685] 1 1 1 1 1 0 1 0 1 0 1 1 1 1 0 0 1 0 0 0 1 0 0 1 0 0 1 1 0 0 1 1 0 1 0 0 1 1
[723] 1 0 0 0 0 1 0 1 0 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 1 1 1 1 1 1 0 1 0 0 0 1 1 1
[761] 0 1 1 0 1 1 1 0 0 0 1 0 1 1 1 1 0 1 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1 1
[799] 1 1 0 0 0 0 1 1 1 1 1 0 0 0 1 1 0 0 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 0 1 0 0 1
[837] 0 0 0 0 1 1 1 1 0 0 1 0 0 0 1 0 0 0 0 1 0 0 0 1 0 1 1 1 0 0 1 1 0 1 1 0 0 1
[875] 0 0 0 1 1 0 0 1 0 1 0 0 1 1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0 0 1 1 1 0 0 0 0 0
[913] 1 0 1 0 1 0 1 1 1 0 1 0 0 0 1 1 0 1 0 0 0 1 1 0 0 1 0 0 1 0 0 0 1 1 0 1 1 0
[951] 0 0 0 0 1 0 1 0 0 1 1 1 1 1 1 0 0 0 1 1 0 1 1 0 1 1 0 1 0 1 1 1 0 1 1 0 0 1
[989] 1 0 0 0 0 1 0 0 1 1 1 1
[ reached getOption("max.print") -- omitted 41613 entries ]
>
RStudio
> summary_by_segment <- male_df %>% group_by(treatment) %>%
+ summarise(conversion_rate = mean(conversion), spend_mean = mean(spend), count = n())
`summarise()` ungrouping output (override with `.groups` argument)
>
RStudio
> summary(summary_by_segment)
treatment conversion_rate spend_mean count
Min. :0.00 Min. :0.005726 Min. :0.6528 Min. :21306
1st Qu.:0.25 1st Qu.:0.007427 1st Qu.:0.8452 1st Qu.:21306
Median :0.50 Median :0.009129 Median :1.0377 Median :21306
Mean :0.50 Mean :0.009129 Mean :1.0377 Mean :21306
3rd Qu.:0.75 3rd Qu.:0.010830 3rd Qu.:1.2302 3rd Qu.:21307
Max. :1.00 Max. :0.012531 Max. :1.4226 Max. :21307
>
> summary_by_segment
# A tibble: 2 x 4
treatment conversion_rate spend_mean count
<dbl> <dbl> <dbl> <int>
1 0 0.00573 0.653 21306
2 1 0.0125 1.42 21307
>
>
####__broomパッケージ__を使うことで、複数本の回帰分析モデルを一括実行することができる。
- purrrとbroomで複数の回帰モデルを効率的に管理する
- (R調べもの) broom::tidyとかtidyMCMCとか (7/26 追記)
- Rを使ったモデル構築の最善策を求めて: {dplyr} + {tidyr} + {broom} + {purrr}を使ったアプローチ
- purrrとbroomで複数の回帰モデルを効率的に管理する
まず最初に、複数本の重回帰モデル式を配列に格納する。
RStudio
> library(broom)
> formulae_vec <- c(spend ~ treatment + recency + channel,
+ spend ~ treatment + recency + channel + history,
+ history ~ treatment + recency + channel)
>
>
> names(formulae_vec) <- paste("reg", LETTERS[1:3], sep="_")
RStudio
> formulae_vec
$reg_A
spend ~ treatment + recency + channel
$reg_B
spend ~ treatment + recency + channel + history
$reg_C
history ~ treatment + recency + channel
>
RStudio
> names(formulae_vec)
[1] "reg_A" "reg_B" "reg_C"
>
配列をDataFrame型に型変換する。
RStudio
> models <- formulae_vec %>% enframe(name = "model_index", value="formula")
RStudio
> models
# A tibble: 3 x 2
model_index formula
<chr> <list>
1 reg_A <formula>
2 reg_B <formula>
3 reg_C <formula>
>
> models$formula
[[1]]
spend ~ treatment + recency + channel
[[2]]
spend ~ treatment + recency + channel + history
[[3]]
history ~ treatment + recency + channel
>
- modelsデータフレームオブジェクト__のformula列__に格納されている3本の重回帰モデルの定義式を、map関数__を使って、lm()関数に渡した上で、male_dfオブジェクト__に対して関数適用する。以上を、__*%>%*パイプ演算子__を使ってパイプライン形式で順番に実行する。
- 出力された結果のオプジェクトを、__$%>%$パイプ演算子__を使って、__tidy()関数__に渡して、データの出力書式を見やすい形に整形する。
RStudio
> df_models <- models %>%
+ mutate(model = map(.x = formula, .f = lm, data = male_df)) %>%
+ mutate(lm_result = map(.x = model, .f = tidy))
>
###結果
RStudio
> df_models
# A tibble: 3 x 4
model_index formula model lm_result
<chr> <list> <list> <list>
1 reg_A <formula> <lm> <tibble [5 × 5]>
2 reg_B <formula> <lm> <tibble [6 × 5]>
3 reg_C <formula> <lm> <tibble [5 × 5]>
>
RStudio
> df_models$lm_result
[[1]]
# A tibble: 5 x 5
term estimate std.error statistic p.value
<chr> <dbl> <dbl> <dbl> <dbl>
1 (Intercept) 1.17 0.242 4.83 0.00000138
2 treatment 0.771 0.145 5.31 0.000000112
3 recency -0.0698 0.0208 -3.35 0.000812
4 channelPhone -0.216 0.237 -0.911 0.362
5 channelWeb -0.0429 0.236 -0.182 0.856
[[2]]
# A tibble: 6 x 5
term estimate std.error statistic p.value
<chr> <dbl> <dbl> <dbl> <dbl>
1 (Intercept) 0.482 0.306 1.57 0.116
2 treatment 0.768 0.145 5.29 0.000000125
3 recency -0.0525 0.0214 -2.46 0.0139
4 channelPhone 0.136 0.256 0.533 0.594
5 channelWeb 0.307 0.255 1.20 0.229
6 history 0.00116 0.000318 3.64 0.000272
[[3]]
# A tibble: 5 x 5
term estimate std.error statistic p.value
<chr> <dbl> <dbl> <dbl> <dbl>
1 (Intercept) 593. 3.69 161. 0
2 treatment 2.72 2.21 1.23 0.220
3 recency -14.9 0.318 -46.9 0
4 channelPhone -304. 3.61 -84.3 0
5 channelWeb -302. 3.60 -83.9 0
>
###unnest関数を使う場合
RStudio
> df_results <- df_models %>%
+ mutate(formula = as.character(formula)) %>%
+ select(formula, model_index, lm_result) %>%
+ unnest(cols = c(lm_result))
>
RStudio
> df_results
# A tibble: 16 x 7
formula model_index term estimate std.error statistic p.value
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 spend ~ treatment + recency + channel reg_A (Intercept) 1.17 0.242 4.83 0.00000138
2 spend ~ treatment + recency + channel reg_A treatment 0.771 0.145 5.31 0.000000112
3 spend ~ treatment + recency + channel reg_A recency -0.0698 0.0208 -3.35 0.000812
4 spend ~ treatment + recency + channel reg_A channelPhone -0.216 0.237 -0.911 0.362
5 spend ~ treatment + recency + channel reg_A channelWeb -0.0429 0.236 -0.182 0.856
6 spend ~ treatment + recency + channel + history reg_B (Intercept) 0.482 0.306 1.57 0.116
7 spend ~ treatment + recency + channel + history reg_B treatment 0.768 0.145 5.29 0.000000125
8 spend ~ treatment + recency + channel + history reg_B recency -0.0525 0.0214 -2.46 0.0139
9 spend ~ treatment + recency + channel + history reg_B channelPhone 0.136 0.256 0.533 0.594
10 spend ~ treatment + recency + channel + history reg_B channelWeb 0.307 0.255 1.20 0.229
11 spend ~ treatment + recency + channel + history reg_B history 0.00116 0.000318 3.64 0.000272
12 history ~ treatment + recency + channel reg_C (Intercept) 593. 3.69 161. 0
13 history ~ treatment + recency + channel reg_C treatment 2.72 2.21 1.23 0.220
14 history ~ treatment + recency + channel reg_C recency -14.9 0.318 -46.9 0
15 history ~ treatment + recency + channel reg_C channelPhone -304. 3.61 -84.3 0
16 history ~ treatment + recency + channel reg_C channelWeb -302. 3.60 -83.9 0
>
RStudio
> df_results %>% filter(model_index == "reg_A")
# A tibble: 5 x 7
formula model_index term estimate std.error statistic p.value
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 spend ~ treatment + recency + channel reg_A (Intercept) 1.17 0.242 4.83 0.00000138
2 spend ~ treatment + recency + channel reg_A treatment 0.771 0.145 5.31 0.000000112
3 spend ~ treatment + recency + channel reg_A recency -0.0698 0.0208 -3.35 0.000812
4 spend ~ treatment + recency + channel reg_A channelPhone -0.216 0.237 -0.911 0.362
5 spend ~ treatment + recency + channel reg_A channelWeb -0.0429 0.236 -0.182 0.856
>
RStudio
> df_results %>% filter(model_index == "reg_B")
# A tibble: 6 x 7
formula model_index term estimate std.error statistic p.value
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 spend ~ treatment + recency + channel + history reg_B (Intercept) 0.482 0.306 1.57 0.116
2 spend ~ treatment + recency + channel + history reg_B treatment 0.768 0.145 5.29 0.000000125
3 spend ~ treatment + recency + channel + history reg_B recency -0.0525 0.0214 -2.46 0.0139
4 spend ~ treatment + recency + channel + history reg_B channelPhone 0.136 0.256 0.533 0.594
5 spend ~ treatment + recency + channel + history reg_B channelWeb 0.307 0.255 1.20 0.229
6 spend ~ treatment + recency + channel + history reg_B history 0.00116 0.000318 3.64 0.000272
>
RStudio
> df_results %>% filter(model_index == "reg_C")
# A tibble: 5 x 7
formula model_index term estimate std.error statistic p.value
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 history ~ treatment + recency + channel reg_C (Intercept) 593. 3.69 161. 0
2 history ~ treatment + recency + channel reg_C treatment 2.72 2.21 1.23 0.220
3 history ~ treatment + recency + channel reg_C recency -14.9 0.318 -46.9 0
4 history ~ treatment + recency + channel reg_C channelPhone -304. 3.61 -84.3 0
5 history ~ treatment + recency + channel reg_C channelWeb -302. 3.60 -83.9 0
>
##ファイル出力
ここからは、「効果検証入門」の本にはありませんが、参考まで。
RStudio
> linear_regression_1 <- lm(spend ~ treatment + recency + channel, data = male_df)
> linear_regression_2 <- lm(spend ~ treatment + recency + channel + history, data = male_df)
####以下を実行すると、Desktop上にhtmlファイルと$LaTeX$ファイルが出力される。
RStudio
> getwd()
[1] "/Users/ocean"
>
> setwd("/Users/ocean/Desktop/")
> getwd()
[1] "/Users/ocean/Desktop"
>
RStudio
> install.packages("stargazer")
URL 'https://cran.rstudio.com/bin/macosx/contrib/4.0/stargazer_5.2.2.tgz' を試しています
Content type 'application/x-gzip' length 619830 bytes (605 KB)
==================================================
downloaded 605 KB
The downloaded binary packages are in
/var/folders/dw/kq_8pwps5s771nklvc8krws40000gn/T//RtmpwaZ17t/downloaded_packages
>
RStudio
> library(stargazer)
Please cite as:
Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.
R package version 5.2.2. https://CRAN.R-project.org/package=stargazer
>
htmlファイルへの書き出し
RStudio
> stargazer(linear_regression_1, linear_regression_2, title = "線形重回帰分析", type="html", out="重回帰分析.html")
<table style="text-align:center"><caption><strong>線形重回帰分析</strong></caption>
<tr><td colspan="3" style="border-bottom: 1px solid black"></td></tr><tr><td style="text-align:left"></td><td colspan="2"><em>Dependent variable:</em></td></tr>
<tr><td></td><td colspan="2" style="border-bottom: 1px solid black"></td></tr>
<tr><td style="text-align:left"></td><td colspan="2">spend</td></tr>
<tr><td style="text-align:left"></td><td>(1)</td><td>(2)</td></tr>
<tr><td colspan="3" style="border-bottom: 1px solid black"></td></tr><tr><td style="text-align:left">treatment</td><td>0.771<sup>***</sup></td><td>0.768<sup>***</sup></td></tr>
<tr><td style="text-align:left"></td><td>(0.145)</td><td>(0.145)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td style="text-align:left">recency</td><td>-0.070<sup>***</sup></td><td>-0.053<sup>**</sup></td></tr>
<tr><td style="text-align:left"></td><td>(0.021)</td><td>(0.021)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td style="text-align:left">channelPhone</td><td>-0.216</td><td>0.136</td></tr>
<tr><td style="text-align:left"></td><td>(0.237)</td><td>(0.256)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td style="text-align:left">channelWeb</td><td>-0.043</td><td>0.307</td></tr>
<tr><td style="text-align:left"></td><td>(0.236)</td><td>(0.255)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td style="text-align:left">history</td><td></td><td>0.001<sup>***</sup></td></tr>
<tr><td style="text-align:left"></td><td></td><td>(0.0003)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td style="text-align:left">Constant</td><td>1.167<sup>***</sup></td><td>0.482</td></tr>
<tr><td style="text-align:left"></td><td>(0.242)</td><td>(0.306)</td></tr>
<tr><td style="text-align:left"></td><td></td><td></td></tr>
<tr><td colspan="3" style="border-bottom: 1px solid black"></td></tr><tr><td style="text-align:left">Observations</td><td>42,613</td><td>42,613</td></tr>
<tr><td style="text-align:left">R<sup>2</sup></td><td>0.001</td><td>0.001</td></tr>
<tr><td style="text-align:left">Adjusted R<sup>2</sup></td><td>0.001</td><td>0.001</td></tr>
<tr><td style="text-align:left">Residual Std. Error</td><td>14.990 (df = 42608)</td><td>14.988 (df = 42607)</td></tr>
<tr><td style="text-align:left">F Statistic</td><td>10.345<sup>***</sup> (df = 4; 42608)</td><td>10.930<sup>***</sup> (df = 5; 42607)</td></tr>
<tr><td colspan="3" style="border-bottom: 1px solid black"></td></tr><tr><td style="text-align:left"><em>Note:</em></td><td colspan="2" style="text-align:right"><sup>*</sup>p<0.1; <sup>**</sup>p<0.05; <sup>***</sup>p<0.01</td></tr>
</table>
>
LaTeXファイルへの書き出し
RStudio
> stargazer(linear_regression_1, linear_regression_2, title = "線形重回帰分析", type="latex", out="重回帰分析.tex")
% Table created by stargazer v.5.2.2 by Marek Hlavac, Harvard University. E-mail: hlavac at fas.harvard.edu
% Date and time: 木, 12 31, 2020 - 17時45分38秒
\begin{table}[!htbp] \centering
\caption{線形重回帰分析}
\label{}
\begin{tabular}{@{\extracolsep{5pt}}lcc}
\\[-1.8ex]\hline
\hline \\[-1.8ex]
& \multicolumn{2}{c}{\textit{Dependent variable:}} \\
\cline{2-3}
\\[-1.8ex] & \multicolumn{2}{c}{spend} \\
\\[-1.8ex] & (1) & (2)\\
\hline \\[-1.8ex]
treatment & 0.771$^{***}$ & 0.768$^{***}$ \\
& (0.145) & (0.145) \\
& & \\
recency & $-$0.070$^{***}$ & $-$0.053$^{**}$ \\
& (0.021) & (0.021) \\
& & \\
channelPhone & $-$0.216 & 0.136 \\
& (0.237) & (0.256) \\
& & \\
channelWeb & $-$0.043 & 0.307 \\
& (0.236) & (0.255) \\
& & \\
history & & 0.001$^{***}$ \\
& & (0.0003) \\
& & \\
Constant & 1.167$^{***}$ & 0.482 \\
& (0.242) & (0.306) \\
& & \\
\hline \\[-1.8ex]
Observations & 42,613 & 42,613 \\
R$^{2}$ & 0.001 & 0.001 \\
Adjusted R$^{2}$ & 0.001 & 0.001 \\
Residual Std. Error & 14.990 (df = 42608) & 14.988 (df = 42607) \\
F Statistic & 10.345$^{***}$ (df = 4; 42608) & 10.930$^{***}$ (df = 5; 42607) \\
\hline
\hline \\[-1.8ex]
\textit{Note:} & \multicolumn{2}{r}{$^{*}$p$<$0.1; $^{**}$p$<$0.05; $^{***}$p$<$0.01} \\
\end{tabular}
\end{table}
Terminalでファイルが生成されていることを確認
Terminal
ocean@AfoGuardMacBook-Pro Desktop % ls | grep tex
mactex.txt
重回帰分析.tex
ocean@AfoGuardMacBook-Pro Desktop % ls | grep html
重回帰分析.html
ocean@AfoGuardMacBook-Pro Desktop %