循环遍历R中的字符串字符答案

【问题标题】：Loop over string characters in R循环遍历R中的字符串字符
【发布时间】：2020-11-24 00:47:08
【问题描述】：

我正在尝试遍历字符列表：

covariate_names <- c("Age_50.", "Age_30.49", "Age_18.29", "Income_High", "Income_Medium", "Income_Low",
                "DemographicSegment_Value.Seeker", "DemographicSegment_Established", "DemographicSegment_Planner", "DemographicSegment_Affluent",
                "DemographicSegment_Digital.Native","Gender_F..", "Gender_M..", "TransactionAmount",
                "TechTxns", "FashionTxns", "TravelTxns", "GamerTxns",
                "Recency", "Frequency", "Monetary", "Breadth", "Consistency", "is_donor", "donation_amt")

pairs <- function(df, covariate, group){
    pwc <- pairwise_t_test(data = df,
                           formula = as.formula(paste0(covariate,"~",group)),
                           paired = FALSE )
    pwc <- as.data.frame(pwc)
    pwc <- pwc %>% rename(GroupA = group1,
                     GroupB = group2,
                     N_grpA = n1, N_grpB = n2)
pwc <- pwc[,1:7]
pwc <- pwc %>%
    mutate_if(is.numeric, round, digits = 6)
    
    #print(pwc)
}

for (i in covariate_names){
    pwc_i <- pairs(df = df_test_TS, covariate = i, group = "w.contextual")
}

但是pairs 函数返回一个df，所以我不知道这是否可以使用循环。我只想为列表中的所有协变量运行pairs函数，并且能够调用pairs函数的每个单独迭代的输出。

【问题讨论】：

标签： r loops for-loop

【解决方案1】：

您可以使用该功能：

library(dplyr)
library(purrr)

pairs <- function(df, covariate, group){
  pwc <- pairwise_t_test(data = df,
                         formula = reformulate(group, covariate),
                         paired = FALSE )
  pwc <- as.data.frame(pwc)
  pwc %>% 
    rename(GroupA = group1,
           GroupB = group2,
           N_grpA = n1, N_grpB = n2) %>%
    select(1:7) %>%
    mutate(across(where(is.numeric), round, digits = 6))
    #For older dplyr
    #mutate_if(is.numeric, round, digits = 6)
  
}

然后使用map/lapply 获取列表作为输出：

result <- map(covariate_names, ~pairs(df_test_TS, .x, "w.contextual"))

如果您想将结果合并到一个数据框中，请使用map_df：

result <- map_df(covariate_names, ~pairs(df_test_TS, .x, "w.contextual"))

【讨论】：

【解决方案2】：

您需要使用循环将结果存储在列表中。您可以尝试下一个代码：

#Data
covariate_names <- c("Age_50.", "Age_30.49", "Age_18.29", "Income_High", "Income_Medium", "Income_Low",
                     "DemographicSegment_Value.Seeker", "DemographicSegment_Established", "DemographicSegment_Planner", "DemographicSegment_Affluent",
                     "DemographicSegment_Digital.Native","Gender_F..", "Gender_M..", "TransactionAmount",
                     "TechTxns", "FashionTxns", "TravelTxns", "GamerTxns",
                     "Recency", "Frequency", "Monetary", "Breadth", "Consistency", "is_donor", "donation_amt")
#Function
pairs <- function(df, covariate, group){
  pwc <- pairwise_t_test(data = df,
                         formula = as.formula(paste0(covariate,"~",group)),
                         paired = FALSE )
  pwc <- as.data.frame(pwc)
  pwc <- pwc %>% rename(GroupA = group1,
                        GroupB = group2,
                        N_grpA = n1, N_grpB = n2)
  pwc <- pwc[,1:7]
  pwc <- pwc %>%
    mutate_if(is.numeric, round, digits = 6)
  
  #print(pwc)
}

这里的变化：

#List to store results
List <- list()
#Loop
for (i in 1:length(covariate_names)){
  List[[i]] <- pairs(df = df_test_TS, covariate = covariate_names[i], group = "w.contextual")
}

【讨论】：