Alternative covariate adjustment procedure.
Exponential and logarithmic fits, with higher-order polynomials coming soon. dta files and also choose to omit the do-file created by savedata() with the nodofile option. Overlay quantiles of the sample distribution on top of the means/medians within each bin, providing more information on the shape of the conditional distribution of y given x.
When working with data involving tens or hundreds of millions of observations, binscatter2 runs between three and eight times faster than binscatter, with the largest relative performance gains on very large datasets. The original binscatter program is extremely well-written and was very efficient when it was written however, recent improvements made possible by the Stata program gtools have allowed several of the operations underlying binscatter to be accomplished much more efficiently, as demonstrated in the benchmarks below. Michael Stepner has provided a wonderful slide deck describing binned scatterplots on his website, available here.Īnyone who has used binscatter on a large dataset can appreciate that it takes a while to run. They are useful for examining the relationship between variables, possibly conditional on a set of covariates and/or fixed effects. Motivationīinned scatterplots are a convenient, non-parametric method to visualize conditional expectation functions. (2019), additional options for fit lines and saving, and multi-way fixed effects. quantile intervals), an alternative procedure to adjust for controls suggested by Cattaneo et al. In addition, binscatter2 offers a handful of new features: the ability to overlay additional information about the conditional probability distribution (e.g. It inherits the syntax and functionality of the excellent binscatter package, but runs substantially faster for big datasets (see benchmarks).
Faster binned scatterplots in Stata with a few new bells and whistlesīinscatter2 is a program for producing binned scatterplots in Stata.
0 kommentar(er)
