Shouldn't you use the "prepared" abcd for matching?

[jacobvjk]

          Shouldn't you use the "prepared" abcd for matching? 

If we are potentially removing inactive companies anyway, it could lead to a slightly easier matching process.

Originally posted by @jdhoffa in #8 (comment)

AB#10475

Advantage would be to slightly improve matching performance and use one consistent ABCD data set throughout the analysis.

Drawback is that the removal of inactive companies is only needed for aggregate metrics, not standard PACTA and applying the prep/removal will add an unnecessary step for standard PACTA potentially reducing the match success rate.