I have data on capital K and profit rates P pertaining to a set of firms X.
S-plus will, of course, easily produce histograms and kernel density
estimates of the distribution of P across X.
However, the theory I am testing makes predictions about the profit rate
distribution weighted by size of capital -- i.e. the probability that a<P<b
is given by sum[Ki/Kt] where the Ki are the amounts of capital employed by
the Xi firms achieving profit rates lying in the interval ab, and Kt is the
total capital employed by all firms.
This is easily worked out and graphed as a histogram in Excel using its PROB
function, but as a beginner I can't see a route to achieving the same result
in S-plus. I'd also like to produce kernel estimates, as well as the
straight histograms.
I take it that the answers aren't likely to depend on the version and
platform involved, but for the record I am using S-plus 2000 under Windows
95.
Julian Wells
OU Business School
The Open University
Walton Hall
Milton Keynes
MK7 6AA
United Kingdom
+44 1908 654658
|