Economics Job Market Rumors » Econometrics Job Rumors

Has anyone used GLLAMM for Stata?

(10 posts)
  • Started 5 months ago by Anonymous
  • Latest reply from anonymous
  1. Anonymous
    Unregistered

    What is your impression/opinion about it?

    Posted 5 months ago #
  2. Anonymous
    Unregistered

    yes. terribly slow and inefficient. SAS does far better work.

    Posted 5 months ago #
  3. Anonymous
    Unregistered

    it is pretty good, versatile

    Posted 5 months ago #
  4. Anonymous
    Unregistered

    I am just trying to estimate a simple multinomial logit with random effects, any recommendations? (using stata)

    Posted 5 months ago #
  5. Anonymous
    Unregistered

    Yes. Stop with this shit and just go OLS.

    Posted 5 months ago #
  6. Anonymous
    Unregistered

    Totally simple in limdep.

    Posted 5 months ago #
  7. Anonymous
    Unregistered

    ^^^ R. Lightning fast at that.

    Posted 5 months ago #
  8. Anonymous
    Unregistered

    Thanks

    Posted 5 months ago #
  9. Anonymous
    Unregistered

    ^^ what do you mean? Its not implemented AFAIK with some guy only having a broken translation of Train code on his site. Consider posting it somewhere if you werent dumb enough to miss the "multinomial" part.

    Posted 2 months ago #
  10. Anonymous
    Unregistered

    The mlogit package in R can estimate MNL with random coefficients. It's pretty fast with small datasets, but a lot slower with larger datasets.

    I've also translated Train's GAUSS code for mixed logits to R and posted it to my website. It's a bit clunky, but it works, and it seems to be faster than mlogit with large datasets. If you experience any problems with it, please send me an email and I'll fix it.

    -Garrett Glasgow

    Posted 8 hours ago #

RSS feed for this topic

Reply

(required)

Allowed markup: a blockquote code em strong ul ol li.
You can also put code in between backtick ( ` ) characters.