relevel.default(delays.df$ORIGIN, ref = "IAD") 中的错误：'relevel' 仅适用于（无序）因素

Question

尝试处理我书中的这段代码（它不断给我不工作的代码）

#### Table 10.6

delays.df <- read.csv("/Users/CHAPTER 10 ASSIGNMENT 6/FlightDelays.csv")
head(delays.df)

# transform variables and create bins

delays.df$DAY_WEEK <- factor(delays.df$DAY_WEEK, levels = c(1:7),
                             labels = c("Mon", "Tue", "Wed", "Thu", "Fri", "Sat", "Sun"))
delays.df$CRS_DEP_TIME <- factor(round(delays.df$CRS_DEP_TIME/100))

head(delays.df)

DAY_WEEK     CRS_DEP_TIME
<fctr>        <fctr>


# create reference categories

delays.df$ORIGIN <- relevel(delays.df$ORIGIN, ref = "IAD")
delays.df$DEST <- relevel(delays.df$DEST, ref = "LGA")
delays.df$CARRIER <- relevel(delays.df$CARRIER, ref = "US")
delays.df$DAY_WEEK <- relevel(delays.df$DAY_WEEK, ref = "Wed")
delays.df$isDelay <- 1 * (delays.df$Flight.Status == "delayed")

# create training and validation sets
selected.var <- c(10, 1, 8, 4, 2, 9, 14)
train.index <- sample(c(1:dim(delays.df)[1]), dim(delays.df)[1]*0.6)  
train.df <- delays.df[train.index, selected.var]
valid.df <- delays.df[-train.index, selected.var]

# run logistic model, and show coefficients and odds
lm.fit <- glm(isDelay ~ ., data = train.df, family = "binomial")
data.frame(summary(lm.fit)$coefficients, odds = exp(coef(lm.fit))) 

round(data.frame(summary(lm.fit)$coefficients, odds = exp(coef(lm.fit))), 5)

由于以下错误，我无法通过任何 relevel 命令：

delays.df$ORIGIN <- relevel(delays.df$ORIGIN, ref = "IAD")

relevel.default(delays.df$ORIGIN, ref = "IAD") 中的错误： 'relevel' 仅适用于（无序的）因素

我的理解是我的变量是因子，relevel 命令不适用于我的因子。我不明白如何解决它？如果不是这种情况，我真的不明白如何解决它。

一如既往，感谢任何帮助或见解。

Answer 1

在 R 4.0 之前，选项 stringsAsFactors 默认为 TRUE，即当被（例如）read.csv() 读入时，类似字符串的变量会自动转换为因子。除非您在读取数据之前设置options(stringsAsFactors=TRUE)（或在您的read.csv() 调用中将该选项指定为参数，即read.csv("[filename]", stringsAsFactors=TRUE)），否则情况不再如此。

您可以通过其中一种方式设置选项，或者您可以在 relevel() 之前显式转换为因子：

delays.df$ORIGIN <- relevel(factor(delays.df$ORIGIN), ref = "IAD")

明智地使用transform()（基本R）或mutate()（tidyverse）可以使您的设置代码更简洁。