R:字符变量必须在.C/.Fortran中重复

Question

我正在尝试使用该leaps包.该数据框df有9列.第2列到第8列是解释变量,第9列是响应变量.df也有(列)名称.

当我尝试使用leaps包时,我得到了一个神秘的错误.

x <- df[,2:8]
y <- df[,9]
leaps <- regsubsets(x, y)
Error in leaps.setup(x, y, wt = weights, nbest = nbest, nvmax = nvmax,  : 
  character variables must be duplicated in .C/.Fortran

这个错误意味着什么,我该如何防止这种情况？

这是data.frame的片段:

> dput(df[1:2,])
structure(list(Var1 = c(2396, 2396), Var2 = c(NA_character_, 
NA_character_), Var3 = c(NA_character_, NA_character_), Var4 = c(501, 
511), Var5 = c(5, 5), Var6 = c(13, 8), Var7 = c(NA_real_, NA_real_
), Var8 = c(NA_real_, NA_real_), Var9 = c(0.0047, 0.0371)), .Names = c("Var1", 
"Var2", "Var3", "Var4", "Var5", "Var6", "Var7", "Var8", "Var9"
), row.names = 1:2, class = "data.frame")

> str(df)
'data.frame':   10000 obs. of  9 variables:
 $ Var1: num  2396 2396 2396 2396 2396 ...
 $ Var2: chr  NA NA NA NA ...
 $ Var3: chr  NA NA NA NA ...
 $ Var4: num  501 511 523 757 770 803 803 803 807 506 ...
 $ Var5: num  5 5 3 5 1 1 5 5 5 5 ...
 $ Var6: num  13 8 13 11 13 8 13 8 11 11 ...
 $ Var7: num  NA NA NA NA NA NA NA NA NA NA ...
 $ Var8: num  NA NA NA NA NA NA NA NA NA NA ...
 $ Var9: num  0.0047 0.0371 0.042 0.0488 0.0048 ...

我尝试将缺少的值替换为0只是为了看看它是否可行,这没有帮助.

Answer 1

我根本不熟悉,leaps但我可以复制你的问题.我认为问题在于其中一个变量是字符或因素:

library("leaps")
df <- data.frame( foo = 1:10, bar = as.factor(1:10), resp = 1:10)
regsubsets(df[,-3],df[,3])

给出错误.请注意,bar这是一个因素:

sapply(df,is.factor)
  foo   bar  resp 
FALSE  TRUE FALSE 

强制转换为数字错误消失了:

df$bar <- as.numeric(as.character(df$bar))
regsubsets(df[,-3],df[,3])

这给出了其他警告,但这可能是由于愚蠢的数据集