Rerun same code but changing colums names at each loop - r

Question

I have a script in r to calculate body condition residuals. I would like to apply this code to each columns, which correspond to a specific category of individual.
For example i would run this code to calculate body condition residuals of all individuals that are in the category 1
1) Select rows of interest
Data1 = RawData %>% select(ID,temperature, Bodysize1, mass1, year) %>% filter((temperature %in% c(20:29) & Bodysize1 %in% c(20:100) & mass1 %in% c(15:40))
2) Create a new model with created data
Model1 =lmer(log(mass1) ~ log(Bodysize1) + temperature + (1|year), data = Data1)
3) Extract residuals and add ID to the residuals
ResModel1 = resid(Model1)
ID=Data1$ID
Res1 =data.frame(ResModel1 ,ID)
4) Add residuals to my RawData
RawData2.0 = merge(RawData, Res1, by = c("ID"), all.x = T)
In order to avoid reruning this code and manually changing all the 1 by 2 and then all the 2 by 3... etc is there a way to do this commande automatic whith loops and the apply familly?
My data
ID TEMPERATURE BODYSIZE1 MASS1 BODYSIZE2 MASS2 YEAR
81-012 0.03830645 200 1450 205 1425 1981
84-069 0.26923078 200 1473 205 1498 1984
84-134 0.32692307 209 1448 195 1323 1984
84-145 0.27884614 197 1373 197 1498 1984
84-190 0.31129807 191 1248 195 1323 1984
85-155 0.33056709 198 1637 229 1988 1985
Thanks in advance

Answer 1

Withou example data it is tough to say if this will work but maybe creating a function can simplify your workflow
library(tidyverse)
get_resid <- function(df,filters) {
df_to_model <- df %>% filter({{filters}})
df_to_keep <- df <- filter({{filters}},.preserve = FALSE)
Data1 <- df_to_model %>%
select(ID,temperature, Bodysize1, mass1, year)
Model1 <- lmer(log(mass1) ~ log(Bodysize1) + temperature + (1|year), data = Data1)
ResModel1 <- resid(Model1)
ID <- Data1$ID
Res1 <- data.frame(ResModel1 ,ID)
Res1 %>%
bind_rows(df_to_keep)
}
Then you may use this this function in your pipes
RawData %>%
get_resid(temperature %in% c(20:29) & Bodysize1 %in% c(20:100) & mass1 %in% c(15:40))

Answer 2

You might try to i) create a tibble, in the first column list all dep variables as strings,
ii) list your models of indep vars in the second column,
iii) create a formula in the third column
iv) run your model in the fourth column
df <- tibble(dep = paste0("log(var",seq(1,10,1),")"),
x = "~ your_x_vars") %>%
mutate(formula = as.formula(paste0(dep,x))) %>%
mutate(reg = map(formula, ~lm(as.formula(.x), data=df) ))
then you can easily extract the residuals