These are public posts tagged with #pymrio. You can interact with them if you have an account anywhere in the fediverse.
New #pymrio release (v0.5.1) with support for #Parquet file storage, GLORIA #mrio downloader (parser coming soon) and some bugfixes. First release with LGPL licence. Available on pypi pypi.org/project/pymrio/ and #condaforge :
https://anaconda.org/conda-forge/pymrio
The new #parquet based file format reduces the save/read time of the full #exiobase system from around 2 min to 20 sec on a typical SSD/laptop (with half the space req).
Will present parts of this at the #iioa conference #opensource special session, Tuesday 27.6 at 1430.
One last thing before the holidays. I plan some home (to #indecol) and licence change (to #lgpl3) for #pymrio and the #countryconverter #coco (both in #Python ) early 2023. Feedback is very much appreciated at
1. Pymrio: https://github.com/konstantinstadler/pymrio/issues/96
2. Coco: https://github.com/konstantinstadler/country_converter/issues/128
This is an effort to emphasis the community aspect of both project and to increase reuse potential.
Dear all, I plan some restructuring of pymrio in early…
github.comMulti-Regional Input-Output Analysis in Python. Contribute…
github.com@fchichorro
Thanks man!
Had the idea to translate parts of #pymrio to #Julia, but after some initial experiments did not find any significant speedup (it calls to blas/lapack for both anyhow).
Still follow the development though - same as for #rust...