如果有任何函数在数据帧上的循环中
如果有任何函数在数据帧上的循环中
提问于 2018-03-19 18:15:37
我在if/else语句中使用字符串的any函数有问题。注意，函数中的打印("A")只是一个例子。如果列包含某些值，则需要执行一系列操作。
随机生成数据
level=c("Strongly Agree", "Agree", "Neither agree or disagree","Disagree", "Strongly disagree",NA)
df <- data.frame(pre_1=as.character(sample(c("Yes","No", NA), 30, replace = T)),
pre_2=as.character(sample(level, 30, replace = T)),
post_1=as.character(sample(level, 30, replace = T)),
post_2=as.character(sample(c("<90%", "0-80%", ">90", NA), 30, replace = T)),
stringsAsFactors=T)选择所需的数据部分("post_")并根据特定列的值打印一条语句。在本例中，我需要为包含特定行值的列打印"A“："Strongly Agree", "Agree", "Neither agree or disagree","Disagree", "Strongly disagree"
select(df, starts_with("post_")) %>%
length() %>%
seq(1,.,1) %>%
for (i in .){
if (any(c("Neither agree or disagree") == (select(df, starts_with("post_"))[i]))){
print ("A")
} else {print ("B")}
} 这就给出了错误
Error in if (any(c("Neither agree or disagree") == (select(df, starts_with("post_"))[i]))) { :
missing value where TRUE/FALSE needed请注意，如果我在这里运行的代码正确工作。
if (any(c("Neither agree or disagree","Agree") == df[3])){print ("A")} else {
print ("B")}感谢你的任何帮助
回答 3
Stack Overflow用户
回答已采纳
发布于 2018-03-20 10:10:20
我使用了@tobiaspk1 1的建议，并使用rowSums对列的某些值进行条件设置。问题是，我希望包含更多的条件(每列的所有因素)，以确保该函数在其他情况下工作(例如，当中间类别缺失时)。
dfplot <- function(df,prefix){
select(df, starts_with(prefix)) %>%
length() %>%
seq(1,.,1) %>%
for (i in .){
if (dummy(as.character(select(data, starts_with(prefix))[[i]])) == FALSE) {
if (colSums(select(df, starts_with(prefix))[i] == "Agree", na.rm = TRUE) > 0){
factor(select(data, starts_with(prefix))[[i]], c("Strongly Agree", "Agree", "Neither agree or disagree","Disagree", "Strongly disagree"),ordered = T ) %>%
data.frame() %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> agreeplot
print(agreeplot)}
else if (colSums(select(df, starts_with(prefix))[i] == "51-75%", na.rm = TRUE) > 0) {
factor(select(data, starts_with(prefix))[[i]], c("1-25%", "26-50%", "51-75%", "75-90%","91-100%"),ordered = T ) %>%
data.frame() %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> numplot
print(numplot)}
else if(colSums(select(df, starts_with(prefix))[i] == "Somewhat too easy", na.rm = TRUE) > 0) {
factor(select(data, starts_with(prefix))[[i]], c("Very easy", "Somewhat too easy", "About right", "Somewhat challenging","Very challenging"),ordered = T ) %>%
data.frame() %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> aboutplot
print(aboutplot)}
else if(colSums(select(df, starts_with(prefix))[i] == "Too slow", na.rm = TRUE) > 0) {
factor(select(data, starts_with(prefix))[[i]], c("Too slow", "Slow", "About right", "Fast","Too fast"),ordered = T ) %>%
data.frame() %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> rightplot
print(rightplot)}
else if(colSums(select(df, starts_with(prefix))[i] == "Between 3 and 4 hours", na.rm = TRUE) > 0) {
factor(select(data, starts_with(prefix))[[i]], c("Less than 2 hours", "Between 2 and 3 hours", "Between 3 and 4 hours", "Between 4 and 5 hours","More than 5 hours"),ordered = T ) %>%
data.frame() %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> hoursplot
print(hoursplot)}
else {data.frame(select(df, starts_with(prefix))[[i]]) %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)")+
ggtitle(names(select(data, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10))-> elseplot
print(elseplot)}}
else {data.frame(select(df, starts_with(prefix))[[i]]) %>%
na.omit() %>%
ggplot(.,aes(x=.)) +
geom_bar(aes(y = (..count..)/sum(..count..)), stat="count") +
geom_text(aes( label =paste(round((..count..)/sum(..count..)*100),"%"), y= (..count..)/sum(..count..)), stat= "count", vjust = -.5)+
scale_y_continuous(labels=percent,limits = c(-0, 1)) +
scale_x_discrete(drop=FALSE) +
ylab("Relative Frequencies (%)") +
ggtitle(names(select(df, starts_with(prefix)))[i]) +
theme_light(base_size = 12) +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
theme(plot.title = element_text(hjust = 0.5,size = 10)) -> dummyplot
print(dummyplot)}
}
}Stack Overflow用户
发布于 2018-03-19 18:34:10
尽可能避免循环。R的强度是矢量计算！
尝试以下几点
results <- character(nrow(df)) # initialise vector
results <- "B" # set B as default
at_least_one <- rowSums(df == "Strongly Agree", na.rm = TRUE) > 0 # find the rows that contain the word searched at least once
results[at_least_one] <- "A" # change those that contain the word to "A"您可以循环遍历您的值，比如“强烈同意”、“同意”和覆盖结果向量！希望这能帮上忙！
Stack Overflow用户
发布于 2018-03-19 20:14:04
用T/F索引到LETTERS的方法有点幼稚: 1.按grepl选择落入模式colptrn中的列；2.将df转到列表中；3.通过列表项进行sapply，并将它们与清单进行比较；4.如果有TRUE，则FALSE + 1 = "A"，如果没有，则TRUE + 1 = "B"；
fu <- function(df, i, colptrn, na.rm = T){
sapply(as.list(df[grepl(colptrn, colnames(df))]),
function(li) LETTERS[1 + !any(i %in% li, na.rm = na.rm)]
)
}
## Test
fu(df, c("Neither agree or disagree", "Agree"), "post_")
post_1 post_2
"A" "B"
fu(df, c("Neither agree or disagree", "Agree"), ".*")
pre_1 pre_2 post_1 post_2
"B" "A" "A" "B"
fu(df, c("Neither agree or disagree", "Agree"), "postpostup")
named list()页面原文内容由Stack Overflow提供。腾讯云小微IT领域专用引擎提供翻译支持
原文链接：
https://stackoverflow.com/questions/49369804
复制相关文章
相似问题
腾讯云开发者
