Things I worked through today:

- Column clustering into graph object representing correlation between variables
- Apple's Developer App - don't even get me started

#genAI #llm #statistics #dataTypes #dataFormats #analystProgrammer #data #python #statsmodels #scikitLearn #rustLang #swift #iOS

www.statsmodels.org/stable/index... Just perfection and to the core! #statsmodels #DataScience

I was delighted today when I, for the first time, used #statsmodels and #patsy to explore some mobility data. Somehow, all the previous times I've had to do #regression analyses, it's been a cumbersome task, but these libraries are just so easy to use and flexible that I am astonished.

Very excited that #statsmodels (the standard #Python library for statistics) accepted a code contribution that makes it 💯 compatible with upcoming #DataMatrix 2.0 as well as other data formats! No more converting to and from pandas DataFrame. github.com/statsmodels/...

#TimeSeries #ETS #Forecasting #AutoETS #HoltWinters #StatsModels #Nixtla #Python #RStats #DemandForecasting #NewChapter

#TimeSeries #ARIMA #Forecasting #DataScience #MachineLearning #Python #RStats #JuliaLang #StatsModels #AutoARIMA #Econometrics #ForecastingModels #ValeryManokhin #NewChapter

Thank you to #Kone & Mai and Tor Nessling Foundations for supporting this work. A quantitative work like this would not be possible without a robust suite of FOSS tools. My thanks to the maintainers of #QGIS, #pandas, #geopandas, #duckdb, #dask, #statsmodels, #jupyter and many more!

Q for #statistics people, #statsmodels users, #stata users:

I'm replicating a Stata analysis in Python's statsmodels. I've replicated the model: both give the exact same coefficients, z-score, etc.

BUT, they forecast differently: first period the same, and then diverge. What's going on here?