wald.test() (from aod package) included as internal function to avoid archivalBug-fix detrend() for handling of vectors in data
Bug-fix detrend() for character vectors of data length in id
predicted option in detrend() returns predicted values
feis() now has option weights to estimate FEIS models with weighted least squares
Allows estimation of conventional FE models by ‘y ~ x | 1’
Rank deficit columns / NA coefs are now dropped from the model (with warning message)
Bug fix: Omit intercept from transformed data if intercept = TRUE
detrend() allows demeaning by ‘slopes = 1’
Minor documentation and vignette updates
Define collinearity test (R:44:1) dynamic based on QR rank on system
New options se.fit and interval for predict.feis(): Compute standard errors and confidence intervals for predicted values.
New function detrend(): Detrends the input data by the predicted values based on the slope parameters within each group.
Bug fix collinearity handling: now compares equality between predicted x and actual x.
Bug fix for vignette: Add correct suggests version
Minor documentation updates
Bug fix feis(), feistest(), bsfeistest(): Vcov / Standard errors are now computed even if some coefficients are NA.
feistest() and bsfeistest(): added terms option to perform Hausman / Chi_sq test on subset of coefficients only.
feis(): added tol option for detecting linear dependencies in slopes.
feis(): available S3 methods extended by coef, deviance, df.residual, fitted, formula, hatvalues, model.matrix, nobs, predict, residuals, sigma, terms, vcov.
Added general info page feisr-package.
Included dplyr bind_rows() for more efficient transformation of large data (depends on dplyr (>= 1.0.0)). Old fall-back code available for dplyr < 1.0.0.
Added vignette for package use.
More efficient use of solve / crossprod
Manual updated, minor corrections.
Added citation for package.
Corrected linkage and definition of S3 methods.