> Does anyone else find that using the Var.calc option (for
> heteroscedasticity consistent std. errors) in Match() (from the
> Matching library) slows down computation of the matching estimator by
> a lot?
The Var.calc option to Match() slows down the function because an
additional loop through the data is required (with vector operations
in it). And this loop, unlike the loop to find the matches
themselves, is not written in C so it is much slower. With this
option, the standard errors do not assume a constant causal effect so
there are a lot of extra calculations to be done.
I guess it would be easy enough to move this loop into C also. It's
on the list....
Cheers,
JS.
======================================Jasjeet S. Sekhon
Associate Professor
Travers Department of Political Science
Survey Research Center
UC Berkeley
http://sekhon.berkeley.edu/
V: 510-642-9974 F: 617-507-5524
======================================
Brian Quinif writes:
> Does anyone else find that using the Var.calc option (for
> heteroscedasticity consistent std. errors) in Match() (from the
> Matching library) slows down computation of the matching estimator by
> a lot?
>
> I don't really understand why when I use this option it slows down so
> much, but for me it does significantly. I want to use the
> heteroscedasticity consistent std. errors in my project, but as long
> as it takes to compute them, I don't know if I will be able to.
>
> Thanks,
>
> BQ
>
>