I'm wondering whether there is a way to truncate the results that Stata 17's new collection system stores by default. For example, collect automatically stores most e-class results after regress. It is straightforward to add to the collected results, but it is not clear to me if it is possible to not collect default results. For example, if I'm running a regression with hundreds of explanatory variables but have only one coefficient of interest, I'd prefer to store just that result and not the whole output. This default storage behavior becomes an issue when collecting from a large number of memory-intensive commands (e.g., say, 50 ivreghdfe regressions with many covariates, as in a output table where each cell represent a separate regression), such that loading a saved stjson collection becomes a difficult and quite slow computational task due to the size of the collection (in the 100's of MB).
Related Posts with Collect System Selectively Store Results
ttest across multiple categoriesI am trying to replicate Table 2 from this paper http://ftp.iza.org/dp5706.pdf, where the difference…
import excel for few defined rows up to the end of spreadsheet for columnsHi, I am trying to import an excel spreadsheet for few defined rows up to the end of spreadsheet fo…
Test whether the difference between two differences is significantHi, I have a dataset with a continous dependent variable VAR and two binary independent variables. …
selection variable by loop 95% samples in stata16Dear Statistical users: For help, I want to randomly select 95% of a sample for logistic regression,…
About code foldingHi,everyone! When I write commands in do-file editor and there is a lot of commands, I always want …
Subscribe to:
Post Comments (Atom)
0 Response to Collect System Selectively Store Results
Post a Comment