如何在新数据集中将因子正确转换为数值

Question

I convert factor to numeric in my dataset as below

library(dplyr) 
df = data.frame(level= c( 'low', 'medium', 'high', 'very high'))

df$level = as.numeric(revalue(df$level, c('low' = 1, 'medium' =2, 'high'= 3, 'very high'=4)))
df

It's ok. The problem arises when I try to apply this rule for new dataset (I traned the model & and want to predict a new data)

newdude = data.frame(level = c( 'high'))
newdude$level = as.numeric(revalue(newdude$level, c('low' = 1, 'medium' =2, 'high'= 3, 'very high'=4)))
Error
The following `from` values were not present in `x`: low, medium, very high 
> newdude
  level
1     1

I get '1' instead of '3' I can not make for ample

newdude$level = as.numeric(revalue(newdude$level, c( 'high'= 3)))

because I can not know in advance what value it will take

How to fix it?

Answer 1

Try instead

newdude = data.frame(level = factor('high', levels = c('low', 'medium', 'high', 'very high')))

newdude$level
[1] high
Levels: low medium high very high
as.numeric(newdude$level)
[1] 3