Error in model.frame.default : variable lengths differ (found for '(weights)')

Asked Jan 31 '21 at 10:39

Active Jan 31 '21 at 11:33

Viewed 524 times

When I am trying to do WLS to correct for heteroskedasticity, I for the following error:

Do you have any suggestions? I noticed that my regression has list length of 13 and the vector for W has 14. I am not sure how to correct for this.

W <- 1 / lm(abs(R40$residuals) ~ R40$fitted.values)$fitted.values^2

R40w <- lm(Market_Cap ~ MarketCap_Lag + Value_traded_ratio + GDP_growthR, data=pdata, weights = W)
summary(R40w)


Error in model.frame.default(formula = Market_Cap ~ MarketCap_Lag + Value_traded_ratio +  : 
  variable lengths differ (found for '(weights)')

edited Jan 31 '21 at 11:33

asked Jan 31 '21 at 10:39

S_Star

I looked for similar problems, but nothing helped. – S_Star Jan 31 '21 at 10:39
can you include the full code, including libraries etc and is this done in R? – StupidWolf Jan 31 '21 at 11:26
can you edit your question? thanks – StupidWolf Jan 31 '21 at 11:32
I just did. Is it okay? Yes, it is in R. – S_Star Jan 31 '21 at 11:33
no way to reproduce your error. we don't have R40 or R40w. my guess is there is NA inside pdata so you end up with different lengths – StupidWolf Jan 31 '21 at 11:37
see https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example – StupidWolf Jan 31 '21 at 11:37
R40 is my regression without WLS, where R40W is with WLS. And also, there is no NA inside my data for sure. – S_Star Jan 31 '21 at 11:40

Error in model.frame.default : variable lengths differ (found for '(weights)')

0 Answers0