It would be good to bundle this into the base Stata image somehow so that users can replace save and use with gzsave and gzuse and benefit from compression.
I've tried compressing some real life dta files and seen saving around 90% so this definitely seems worthwhile.
The cohortextractor supports writing files as
.dta.gz
but these can't be read natively by Stata without the gzsave package: https://ideas.repec.org/c/boc/bocode/s446701.html https://github.com/ebardelli/gzsaveIt would be good to bundle this into the base Stata image somehow so that users can replace
save
anduse
withgzsave
andgzuse
and benefit from compression.I've tried compressing some real life
dta
files and seen saving around 90% so this definitely seems worthwhile.