Open dewittpe opened 8 years ago
I've come to the conclusion that the function propensity
is too generic of a name and does too much. I'm going to build several functions that do different parts of the work.
dstats
for descriptive statistics from a regression model.standardized_differences
which will take either a regression model or a pstools_dstats
object as input and return standardized differences for each variable by the different weighting methods.summary
and print
methods will also be provided.
The function
psweights()
(issue #5) calculates five different types of propensity matching weights. Use this function within thepropensity()
function and extend thesummary
andplot
methods accordingly.