Mar 9, 2014.   1.7-1286 -- On CRAN.

Mar 5, 2014.   1.7 -- Depend on Matrix >= 1.1-2 to ensure that exactDOF='rM' works.
     Also import xtable and Formula.  Suggest stargazer.

Feb 12, 2014.  1.7 -- Give warnings for unnamed 'clustervar' and 'iv'
     args, in preparation for moving them out of the argument list.

Feb 11, 2014.  1.7 -- Support for multi part formulas.  G() removed from documentation.

Feb 3, 2014.   1.7 -- Do multiway clustering bootstrap also.

Jan 29, 2014.  1.7 -- Support multiway clustering (Cameron, Gelbach,
     Miller). Suggested by Thiemo Fetzer.

Jan 23, 2014.  1.6-1188 -- On CRAN. Map G() to factor(), not
     as.factor(), so as to remove unused levels which otherwise will
     mess up the computation of DOF.

Jan 4, 2014.  1.6-1180 -- On CRAN.

Dec 17, 2013. 1.6 -- Fixed a bug in computing F-test for IV. Added
     F-test in 1st stage for omitted i.v.  Suppress F-test writing for
     robust and clustered errors. I can't compute it.

Dec 06, 2013. 1.6-1163 -- On CRAN.

Nov 29, 2013. 1.6 -- Support for G(x:f) with x matrix.
     Fixed a one-off error in df for the one-factor case.
     Several bug-fixes and memory optimization pertaining to large model matrices.
     Some of them in R.

Nov 25, 2013. 1.6 -- Matrix::rankMatrix bug has been fixed. We can
     use the sparse method qrLINPACK as soon as a new version of
     Matrix is released.  Bug #5031:
     https://r-forge.r-project.org/tracker/index.php?func=detail&aid=5031&group_id=61&atid=294
     Added contrasts-argument to felm. It applies to factors outside
     G().  Fixed efactory() for G(x:f) interactions.

Nov 14, 2013. 1.6 -- Added G(x:f) syntax for estimation similar to
     x:f, i.e. interaction between a covariate and a factor. felm()
     and getfe() are ok, but efactory() does not always produce
     estimable functions.

Nov 12, 2013. 1.5-1118 -- On CRAN. Fixed a bug which prevented bootstrapping in some cases.

Oct 28, 2013. 1.5-1107 -- On CRAN. Version presented in R journal article.
