分裂功能后在r中建立列车模型

Question

我正在构建一个逻辑回归来预测最有效的买家，我有一个数据表，我能够编写代码直到拆分功能，但是数据表中的一列只有$ 100- $ 200，$ 200- $ 300 。如何分配随机数值并继续进行，下面是参考代码

file_data_0 = read.csv('......csv')

View(file_data_0)

file_data_0[file_data_0 == '?'] = NA

filter_data_1 = na.omit(file_data_0)

View(filter_data_1)

summary(filter_data_1)

unique(filter_data_1$spend)

table(filter_data_1$spend)

table(filter_data_1$spend)/nrow(filter_data_1)

class(filter_data_1)

split the data into training and testing

library(caTools)

set.seed(64000)

split_data <- sample.split(filter_data_1$spend, SplitRatio = 0.75)

train_data <- subset(filter_data_1, split_data == TRUE)

test_data <- subset(filter_data_1, split_data == FALSE)

dim(train_data)

dim(test_data)

dim(filter_data_1)

class(filter_data_1$history_segment)

filter_data_1$history_segment = as.numeric(filter_data_1$history_segment)

样本数据表

enter image description here

Answer 1

希望这可以帮助！

#sample data
#I copied it from the attached image file but in future you should always share your data using 'dput(df)'
df <- data.frame(recency = c(10,6,7,9),
                 history_segment=c('2) $100 - $200', '3) $200 - $350', '2) $100 - $200', '5) $500 - $750'),
                 mens=c(1,1,0,1))

#little bit of data cleaning in 'history_segment' column
# otherwise accessing these variables for further processing would be another pain!
df$history_segment <- paste0("his_seg_", gsub("^\\d)|\\s|\\$", "", gsub("-", "_", df$history_segment)))

#dummy code 'history_segment' variable
library(psych)
df_new <- cbind(df[,names(df) != "history_segment"], dummy.code(df$history_segment))

现在你可以使用df_new作为主要数据而不是df

> df
  recency history_segment mens
1      10 his_seg_100_200    1
2       6 his_seg_200_350    1
3       7 his_seg_100_200    0
4       9 his_seg_500_750    1
> df_new
  recency mens his_seg_100_200 his_seg_200_350 his_seg_500_750
1      10    1               1               0               0
2       6    1               0               1               0
3       7    0               1               0               0
4       9    1               0               0               1