CODEWITHSUNDEEP
r
asalasa209
asalasa209

Reputation: 47

How to append coefficients from multiple lm objects into a data frame?

I have a list of 192 data frames, each data frame contains several variables. My objective is to run a regression for each data frame in the list and extract the intercept and the betas and append them into a new data frame, so I will have a data frame with 192 rows and one column with intercepts and several columns for betas.

The only approach I could run was the following:

    # list with 192 lm objects #
    reg_list = list() 

    # loop for running regressions and adding the results into the list #
    for(i in 1:length(df)){  
    reg_list[[i]]= lm(x1 ~ x2 + x3 ,data=df[[i]])
    }

# "df" is the list with the 192 data frames #

The issue here is that I have to extract the coefficients from this new list with another procedure, so I am looking a way to append the coefficients into a data frame directly in the same loop.

Thanks in advance for the help

Upvotes: 0

Views: 535

Answers (1)

G. Grothendieck
G. Grothendieck

Reputation: 269586

This performs the regressions giving list reg and then creates a matrix of coefficients with the ith column holding the ith set of coefficients. anscombe is a data frame that comes with R that has x1, x2, x3 and others as column names.

dfs <- list(anscombe, anscombe) # list of data frames
reg <- lapply(dfs, lm, formula = x1 ~ x2 + x3)
sapply(reg, coef)

Upvotes: 3

Related Questions