Inequality.jl has functions for the analysis of inequality and poverty indicators. It currently supports the estimation of the following (both weighted and unweighted):

- Gini Coefficient
- Atkinson Coefficient
- Lorenz Curve
- Generalized Entropy Index
- Theil Index
- Mean log deviation
- Watts Poverty Index
- Foster–Greer–Thorbecke Index
- Poverty Gap
- Headcount Ratio

It’s my first Julia package so coments, suggestions and contributions are more than welcome!

21 Likes

This looks like a very useful package to have for people working in those areas that’s well documented too!

I do feel that the name could have been chosen to be slightly more specific, such as `InequalityIndicators.jl`

as it is more precise description of what the package provides and fits better with Julia’s package naming conventions.

I apologise that this is my only specific comment on the package (but perhaps this is a compliment to your documentation and concise code!).

21 Likes

Thanks very much for your feedback, Jacobus!

I’m planning to expand the package beyond the current indicators. Most likely including top-income adjustments and functions for processing income and wealth variables (e.g. equivalisations and CPI/PPP adjustments). Thus, I’ll probably keep the current general package name.

Hi,

I published PovertyAndInequalityMeasures.jl a while back, which covers much the same ground. Be interesting to see how they compare. Let a thousand flowers bloom and all that.

Graham

2 Likes

Thank you for the contribution.

The name means something very different to many Julia users,

“inequality” is of a field of mathematics and some computer science.

The orthographic root may be the same; the manner of consideration differs. Speaking for myself, I expected a package that provided tools for investigating or establishing inequations (a = b + c, an equation … replace `=`

with another comparative relationship and there are many uses for the sort of expressions that result, see Inequality of arithmetic and geometric means - Wikipedia). A name like that suggested above `InequalityIndictors`

or `EconomicInequalities`

or what you find more explanatory would help.

21 Likes

Thanks for your feedback, Jeffrey.

I now see why the package name might lead to confusion. I’ll seriously think about changing it.

5 Likes

Hi Graham.

I checked for other packages in the area before writing mine, but I completely missed yours. It’s great to see there’s other people working on inequalities/poverty using Julia!

1 Like

Yes, when I saw the package name, I expected a package for solving mathematical inequalities.

Yes, it’s good to see someone interested in the same stuff. It would be great to see a suite of survey-data related things in Julia, rather like R has. My wishlist includes:

- survey data weighting: I’ve got an implementation that does the job but could do with a lot of development and testing;
- data matching. Like the R Statmatch package;
- multidimensional poverty and inequality;
- a nice tabulator, kind of like SPSS has.

You also mention equivalence scales in your bio. I have a bunch of these in a sub-package of my tax-benefit model. It’s inelegant code, but might be a start.

I’ll have a good look at your package. Looks more elegant than mine …

Graham