R ifelse statement ignores input-CodePudding

I have a problem making an ifelse / if_else statement. I want to create a new column "Transtyp" which is assigned a "2" if the column "Plot" has an uneven number (1 or 3 or 5 or 7) and is assigned a "5" when "Plot" has an even number. This is my df

df <- structure(list(Soll = c("1189", "1189", "1189", "1189", "1189", 
"1189", "1189", "1189", "1189", "1189", "1189", "1189", "1189", 
"1189", "1189", "1189", "1189", "1189", "1189", "1189"), Datum = c("2021-03-15", 
"2021-03-15", "2021-03-15", "2021-03-15", "2021-03-15", "2021-03-16", 
"2021-03-16", "2021-03-16", "2021-03-16", "2021-03-16", "2021-03-17", 
"2021-03-17", "2021-03-17", "2021-03-17", "2021-03-17", "2021-03-18", 
"2021-03-18", "2021-03-18", "2021-03-18", "2021-03-18"), Plot = c(1, 
4, 6, 7, 8, 1, 4, 6, 7, 8, 1, 4, 6, 7, 8, 1, 4, 6, 7, 8), Temp = c(21.4609852941176, 
21.5064705882353, 21.4601323529412, 21.4587352941176, 21.4630147058824, 
11.0828229166667, 11.03209375, 11.3093020833333, 11.2860833333333, 
11.1405104166667, 2.1683125, 2.10675, 2.87871875, 2.84852083333333, 
2.44628125, 1.41326041666667, 1.47109375, 1.66685416666667, 1.53930208333333, 
1.34465625), RH = c(37.8671617647059, 37.8925, 38.0004411764706, 
38.26425, 37.9111029411765, 62.1298645833333, 63.2143333333333, 
62.1141666666667, 62.55275, 62.5445416666667, 92.7891041666667, 
93.5631458333333, 91.8378333333333, 93.2626145833333, 92.565625, 
92.7335625, 93.5223125, 92.7023541666667, 93.6509375, 93.120375
), Transtyp = c(2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 
2, 2, 2, 2, 2)), row.names = c(NA, -20L), class = c("tbl_df", 
"tbl", "data.frame"))

Now, I am using this code, but it only assigns "2" if Plot == 1 and ignores the rest. What am I missing?

df %>% mutate(Plot = as.numeric(Plot)) %>%
        mutate(Transtyp = ifelse(Plot == (1 | 3 | 5 | 7), "2", "5")) -> df_t

Any help is appreciated! Thanks a lot!

CodePudding user response：

You want:

library(dplyr)

df %>% mutate(Transtyp = if_else(Plot %in% c(1, 3, 5, 7), "2", "5"))

# A tibble: 20 × 6
   Soll  Datum       Plot  Temp    RH Transtyp
   <chr> <chr>      <dbl> <dbl> <dbl> <chr>   
 1 1189  2021-03-15     1 21.5   37.9 2       
 2 1189  2021-03-15     4 21.5   37.9 5       
 3 1189  2021-03-15     6 21.5   38.0 5       
 4 1189  2021-03-15     7 21.5   38.3 2       
 5 1189  2021-03-15     8 21.5   37.9 5       
 6 1189  2021-03-16     1 11.1   62.1 2       
 7 1189  2021-03-16     4 11.0   63.2 5       
 8 1189  2021-03-16     6 11.3   62.1 5       
 9 1189  2021-03-16     7 11.3   62.6 2       
10 1189  2021-03-16     8 11.1   62.5 5       
11 1189  2021-03-17     1  2.17  92.8 2       
12 1189  2021-03-17     4  2.11  93.6 5       
13 1189  2021-03-17     6  2.88  91.8 5       
14 1189  2021-03-17     7  2.85  93.3 2       
15 1189  2021-03-17     8  2.45  92.6 5       
16 1189  2021-03-18     1  1.41  92.7 2       
17 1189  2021-03-18     4  1.47  93.5 5       
18 1189  2021-03-18     6  1.67  92.7 5       
19 1189  2021-03-18     7  1.54  93.7 2       
20 1189  2021-03-18     8  1.34  93.1 5

In R, | refers to logical or, so when you do 1 | 3 | 5 | 7 it returns TRUE. Then, when you try to compare Plot to TRUE, it will implicitly convert TRUE to 1. This is why it wasn't behaving as you wanted:

> as.numeric(1 | 3 | 5 | 7)
[1] 1