NEWS | R Documentation |

lag.formula: Bug introduced from previous update which could lead to wrong results. Now fixed.

[All estimation methods] Significant speed improvement when the fixed-effects variables (i.e. the identifiers) are string vectors.

[did_means] New function

`did_means`

to conveniently compare means of groups of observations (both treat/control and pre/post). Contains tools to easily export in Latex.

[All estimation methods] Significant speed improvement when the fixed-effects variables (i.e. the identifiers) are of type integer or double.

[esttex, esttable] New argument 'fitstat' to select which fit statistic to display. The default adapts to the models. Old arguments (loglik, bic, aic, sq.cor) are dropped.

[esttable] Significantly better rendering of SE types.

[r2] Now NA is returned for R2s that have no theoretical justification (e.g. within R2 when no FEs, or 'regular' R2 for ML models).

[did_plot_yearly_effects] Now the name of the dependent variable appears on the y-axis.

[esttex] Usage of the

`sym`

macro in Latex is dropped.

[fixef.fixest] bug could appear when using varying slopes coefficients in some specific circumstances (when the slope FEs were different from the regular FEs).

[fixef.fixest] bug when many regular FEs jointly with varying slopes.

[fixef.fixest] regarding slope coefficients: now the algorithm also evaluates functions of variables.

[esttable] Width of the "separating lines" now appropriately set for long dependent variable names.

[esttex] Spelling mistake corrected.

[estimations] Bug could occur for extremely small data sets (< 10 observations).

[esttex, esttable] More informative error messages in functions esttex and esttable.

lag.formula: When the data was not in a particular format, the results could be wrong. Now corrected.

feglm: bug when a) the deviance at initialization was higher than the deviance of the first iteration of the IRWLS and b) the step-halving was unable to find a lower deviance. This led the estimation to fail with an error although it should have been performed properly.

did_estimate_yearly_effects: bug when the estimation involved periods with negative values

esttex: bug regarding the number of digits of negative coefficients to be displayed

esttex: now properly escaping the percentage and the underscore for exports in Latex

esttex: bug when changing the names of the dependent variables using a dictionnary

vcov: some warning messages were misleading

update: bug update when using the argument nframes

update: bug when updating the function fepois

Better error messages for: did_estimate_yearly_effects, main estimation functions, setFixest_dict, fepois and fenegbin.

This package is an effort to create a family of fast and user-friendly functions to perform estimations with multiple fixed-effects (F.E.).

Estimations with fixed-effects (or call it factor variables) is a staple in social science. Hence having a package gathering many methods with fast execution time is of prime importance. At the time of this version, this is the fastest existing method to perform F.E. estimations (often by orders of magnitude, compared to the most efficient alternative methods [both in R and Stata]). The underlying method to obtain the F.E. is based on Berge 2018, and the workhorse of the code is in c++ parallelized via OpenMP (btw thanks Rcpp for simplifying coders' life!).

This package is the follow up of the (now deprecated) package FENmlm which performed fixed-effects estimations but for only four likelihood families. Package fixest completely supersedes FENmlm by extending the method to regular OLS and all GLM families, and adding new utility functions. Further, the design of the functions has been completely overhauled and extended towards much more user-friendliness. Massive effort has been put into providing a set of informative error messages to the user for quick debugging of her workflow (e.g. one of the functions contains over 100 different errors messages).