CODEWITHSUNDEEP
rdataframelm
Rfanatic
Rfanatic

Reputation: 2280

Regression in R on panel data

I would like to run a linear regression on panel data. Below my code so far but, I don't understand why is not returning the fit and rsq. Any suggestion?

Sample code:

for(i in names(df))
{ 
  if(is.numeric(df[3,i]))  ##if row 3 is numeric, the entire column is 
  {       
    fit <- lm(df[3,i] ~ Gender, data=df) #does a regression for each column in my csv file against my independent variable 'etch'
    rsq <- summary(fit)$r.squared
  }
}

Data structure

enter image description here

Sample data:

    df<-structure(list(id = c(1, 1, 2, 2, 2), id1 = c(1, 2, 1, 2, 3), 
    a1 = c(5, 8, 7, 6, 3), a2 = c(1, 4, 3, 10, 5), a3 = c(2, 
    34, 3, 12, 6), a4 = c(9, 2, 3, 12, 7), a5 = c(0, 0, 0, 7, 
    8), a6 = c(7, 7, 0, 0, 9), a7 = c(5, 8, 7, 6, 0), a8 = c(1, 
    4, 3, 10, 3), a9 = c(2, 34, 3, 12, 3), a10 = c(9, 2, 3, 12, 
    3), Gender = c(1, 2, 1, 1, 2)), class = c("spec_tbl_df", 
"tbl_df", "tbl", "data.frame"), row.names = c(NA, -5L), spec = structure(list(
    cols = list(id = structure(list(), class = c("collector_double", 
    "collector")), id1 = structure(list(), class = c("collector_double", 
    "collector")), a1 = structure(list(), class = c("collector_double", 
    "collector")), a2 = structure(list(), class = c("collector_double", 
    "collector")), a3 = structure(list(), class = c("collector_double", 
    "collector")), a4 = structure(list(), class = c("collector_double", 
    "collector")), a5 = structure(list(), class = c("collector_double", 
    "collector")), a6 = structure(list(), class = c("collector_double", 
    "collector")), a7 = structure(list(), class = c("collector_double", 
    "collector")), a8 = structure(list(), class = c("collector_double", 
    "collector")), a9 = structure(list(), class = c("collector_double", 
    "collector")), a10 = structure(list(), class = c("collector_double", 
    "collector")), Gender = structure(list(), class = c("collector_double", 
    "collector"))), default = structure(list(), class = c("collector_guess", 
    "collector")), skip = 1L), class = "col_spec"))

Upvotes: 0

Views: 110

Answers (1)

Ronak Shah
Ronak Shah

Reputation: 388807

To fit a linear regression on each column you can use lapply. We use reformulate to create a formula object from the column name and use it in lapply. R-squared value can be extracted from summary of each model.

cols <- grep('a\\d+', names(df), value = TRUE)
cols
#[1] "a1"  "a2"  "a3"  "a4"  "a5"  "a6"  "a7"  "a8"  "a9"  "a10"

lapply(cols, function(x) {
  lm(reformulate('Gender', x), df)
}) -> fit

r.squared <- sapply(fit, function(x) summary(x)$r.squared)

Upvotes: 1

Related Questions