Nonlinear least squares with many coefficients

[Dennis]

I am analyzing a nonlinear least squares model with 50,000 observations. I have estimated a model that includes 200 fixed effects as well as about 10 coefficients that enter nonlinearly. I want to extend this model to include approximately 1,000 fixed effects. I wrote the code to do this, but it appears to go beyond Eviews capacity limits. Is there a way around this?

[EViews Gareth]

I assume you mean you have a panel with 50,000 x 1,000 dimensions? That is way too big for the current version of EViews, and there is little way around it.

[EViews Glenn]

I'm assuming that the effects enter nonlinearly in your spec?