[ANN] Announcing ItPropFit.jl

I am happy to announce my first Julia package, ItPropFit.jl! It is a package for multidimensional iterative proportional fitting. It has similar features to the R package mipfp: the target array can be multidimensional, and the corresponding array margins can themselves be multidimensional as well.

Of course, this is a Julia package so we get a nice speed improvement over mipfp (approximately 5 - 20 times speedup depending on the situation on my computer).

I wrote this package based on this thread.

It has full documentation, with an example where I integrate ItPropFit.jl with FreqTables.jl and DataFrames.jl to compute survey weights for a linear regression (see here).

This is my first Julia package, so I am excited to hear what you think. In case you have comments on the code, I am happy to hear from you! Contributions are also very welcome, of course.

For a rough development to-do list, you can check out the github issues.

7 Likes

Congratulations on the package! Consider giving it a more verbose name, that is common practice in Julia. Maybe IterativeProportionalFitting.jl or just ProportionalFitting.jl?

We try to avoid R’s “soup of letters in package names” to help potential users find your work more easily.

4 Likes

Thanks for the feedback. Will do, for sure. I really like ProportionalFitting.jl!

4 Likes