Cristian Arteaga

Hello @alvarogutyerrez, I just wanted to let you know that the latest version of xlogit allows you to specify an additive term that helps to model coefficients kept at a...

Hi @chevrotin, you can pass initial values to xlogit using the `init_coeff` parameter to the `fit` function (see example below). You just need to make sure to pass your initial...

I think the batch processing functionality in xlogit can fix your GPU OutOfMemoryError. Simply pass the `batch_size=1000` parameter to the `fit` method to tell **xlogit to process 1,000 random draws...

Sure, please let me know how it works. Unfortunately xlogit's documentation does not properly elaborate on the batch processing and WTP space models. I definitely need to spend some time...

That would be actually fantastic and I would love your help. If you can create some examples for WTP space models in a Google Colab notebook, I can directly integrate...

Thanks a lot @chevrotin!

Hi @chevrotin, I am glad that the batch processing functionality worked for you and you managed to estimate a model with 50k random draws. Thank you so much for creating...

Thanks a lot for your detailed explanation, @jhelvy. I will check the heuristics included in your multi-start loop. Given the importance of the multi-start functionality, I might include it in...

Hi @chevrotin, you are correct, in xlogit the scale parameter is fixed and does not have an associated standard deviation. Yes, xlogit uses the negative of the scale parameter, as...

Dear @chevrotin, I finally had the chance to integrate into the docs the example you kindly created (see link below): https://xlogit.readthedocs.io/en/latest/notebooks/wtp_space.html Thank you again for the time you dedicated to...