Agfdhyk

I've tried to look for a similar question but I couldn't find anything similar.

I have a data frame df with hundreds of rows and several variables. The first variable is level which goes from 1 to 8.

Eg:

df<- data.frame(level = c(1,1,1,2,2,3,4,5,6,7,8), CODE = c("1234", "3452", "1234", "7654","6547","6546", "7683", "6543", "7683", "6543", "7683"), ADD_ALLOW_MEAL = c(NA, "Y", "Y", "N", "N", NA, NA, "Y", "Y", "N", 'N'), ALLOW_MEALLOW = c(NA, 40, 60, NA, NA, NA, NA, 50, 70, NA, NA)) 



> df

   level     CODE ADD_ALLOW_MEAL     ALLOW_MEALLOW

1      1     1234           <NA>                NA

2      1     3452              Y                40

3      1     1234              Y                60

4      2     7654              N                NA

5      2     6547              N                NA

6      3     6546           <NA>                NA

7      4     7683           <NA>                NA

8      5     6543              Y                50

9      6     7683              Y                70

10     7     6543              N                NA

11     8     7683              N                NA

What I need is to create a new data frame which will have only 8 rows (8 levels from df). Normally I'd use simple:

df %>% 

  group_by(level) %>% 

  summarise()

The problem is I need to create several very custom columns on the filtered data, all per level.

Example:

df %>% 

  group_by(level) %>% 

  summarise(

    Meal_Average =  filter(., ADD_ALLOW_MEAL =="Y" & ALLOW_MEALLOW>0) %>% {ifelse(str_detect(.$CODE, "2")=="TRUE", round(mean(.$ALLOW_MEALLOW, na.rm = TRUE),3), NA_real_ )}  

  )

I get an error of:

Column `Meal_Average` must be length 1 (a summary value), not 4

My desired result would be:

 level    Meal_Average 

1      1   50

2      2   NA

3      3   NA

4      4   NA

5      5   NA

6      6   NA

7      7   NA

8      8   NA

Any ideas how can I do that?

Thanks!!

Here is an idea using dplyr. What I did here is I just used your conditions to replace the ALLOW_MEALLOW value to 0 (so that it does not impact the mean), i.e.

library(dplyr)



df %>% 

 mutate(ALLOW_MEALLOW = replace(ALLOW_MEALLOW, ADD_ALLOW_MEAL == 'N' & ALLOW_MEALLOW < 0 | !grepl('2', CODE), 0)) %>% 

 group_by(level) %>% 

 summarise(new_mean = mean(ALLOW_MEALLOW, na.rm = TRUE))

which gives,

# A tibble: 8 x 2

  level new_mean

  <dbl>    <dbl>

1     1       50

2     2        0

3     3        0

4     4        0

5     5        0

6     6        0

7     7        0

8     8        0

NOTE: You can replace 0 with NA as usual