1

I am trying to save the summary outputs of a lm into a data frame, "csv" or "txt" file. What I wanted to do is to use a lm (and eventually glm) with different dependent variables, but the same independent variables.

This is my code using lapply for model fitting:

varlist <- names(NDVI)[2:244]

models <- lapply(varlist, function(x) {
lm(substitute(i ~ efectohuracan, list(i = as.name(x))), data = NDVI)})

If I save it as 

write.csv(models,"models.csv")

I get this error

Error in as.data.frame.default(x[[i]], optional = TRUE, stringsAsFactors = 
stringsAsFactors) : cannot coerce class ""lm"" to a data.frame

I am very new in R. The code I made it thanks to people who has posted their problems also in this community, so maybe there are better ways to approach this.

Thank you in advance

Zheyuan Li
  • 71,365
  • 17
  • 180
  • 248
Ana Patricia Ruiz Beltrán
  • 9
  • 1
  • 5

1 Answers1

1

Sorry I wrongly closed this as a dupe at the beginning. Thanks to my friend for reopening it.

For "mlm" model class this is very efficient.

First, you need to Fitting a linear model with multiple LHS

Then let fit be your fitted model object (of "mlm" and "lm" class), extract its coefficients by

beta <- coef(summary(fit))

This is a list of coefficient tables.

Then, let's collapse it into a data frame:

tab <- do.call(rbind.data.frame, beta)

Now you just write this data frame into a "csv" file as usual.

As a quick test, you can use the toy dataset provided in the linked thread.

Zheyuan Li
  • 71,365
  • 17
  • 180
  • 248