You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
79 lines
2.7 KiB
Plaintext
79 lines
2.7 KiB
Plaintext
10 months ago
|
.-
|
||
|
|
help for ^prcounts^ - 11Mar2005
|
||
|
|
.-
|
||
|
|
|
||
|
|
Compute the predicted rate and probabilities for count models
|
||
|
|
-------------------------------------------------------------
|
||
|
|
|
||
|
|
^prcounts^ varname [^if^ exp] [^in^ range] [^, m^ax^(^max^)^ ^p^lot]
|
||
|
|
|
||
|
|
Description
|
||
|
|
-----------
|
||
|
|
|
||
|
|
^prcounts^ computes the predicted rate and probabilities of counts from 0
|
||
|
|
through the specified maximum count based on the last estimates from the
|
||
|
|
count models ^poisson^, ^nbreg^, ^zip^, ^zinb^. ^prcounts^ optionally generates
|
||
|
|
variables for the graphical comparison of observed and expected counts
|
||
|
|
for one model or for competing models.
|
||
|
|
|
||
|
|
For each observation ^prcounts^ computes the variables:
|
||
|
|
|
||
|
|
namerate = predicted count or rate
|
||
|
|
namepr^k^ = probability of a count of ^k^ for ^k^=0 to max
|
||
|
|
namecu^k^ = cumulative probabiliy of a count <=^k^ for ^k^=0 to max
|
||
|
|
|
||
|
|
For zip or zinb models, prcounts also generates:
|
||
|
|
|
||
|
|
namell0 = probability of being in the 'always zero' group
|
||
|
|
|
||
|
|
For ztp or ztnb models, prcounts also generates:
|
||
|
|
|
||
|
|
nameCrate = predicted count or rate conditional on y>0
|
||
|
|
nameCpr^k^ = probability of a count of ^k^ for ^k^=0 to max conditional
|
||
|
|
on y>0
|
||
|
|
nameCcu^k^ = cumulative probabiliy of a count <=^k^ for ^k^=0 to max
|
||
|
|
conditional on y>0
|
||
|
|
|
||
|
|
Options
|
||
|
|
-------
|
||
|
|
|
||
|
|
^max(^value^)^ The maximum count for which variables are generated.
|
||
|
|
value must be an integer between 0 and 99; the default is 9.
|
||
|
|
|
||
|
|
^plot^ causes Stata to generate extra variables that are useful for plotting.
|
||
|
|
The created variables will have values for the first max+1 rows of the
|
||
|
|
data set.
|
||
|
|
|
||
|
|
nameval = value of y
|
||
|
|
nameobeq = observed probability of counts = y
|
||
|
|
nameoble = cumulative observed probability of counts <= y
|
||
|
|
namepreq = predicted probability of counts = y
|
||
|
|
nameprle = predicted cumulative probability of counts <= y
|
||
|
|
|
||
|
|
For these variables, each observation provides information about the
|
||
|
|
probabilities of a given count for the entire sample. In other words, what
|
||
|
|
an observation consists of for these variables is different than for the rest
|
||
|
|
of the variables in one's dataset.
|
||
|
|
|
||
|
|
Note: the summary statistics computed by ^plot^ are automatically based on
|
||
|
|
the estimation sample, plus any further restrictions implied by the use of
|
||
|
|
^if^ or ^in^.
|
||
|
|
|
||
|
|
Examples
|
||
|
|
--------
|
||
|
|
|
||
|
|
. ^zinb art fem mar kid5 phd ment, inf(fem mar kid5 phd ment)^
|
||
|
|
. ^prcounts cnt^
|
||
|
|
. ^summarize cnt*^
|
||
|
|
|
||
|
|
To graphically compare predicted and observed counts for a model:
|
||
|
|
|
||
|
|
.^ poisson art fem mar kid5 phd ment^
|
||
|
|
.^ prcounts pois, max(8) plot^
|
||
|
|
.^ graph poispreq poisobeq poisval, c(ll)^
|
||
|
|
|
||
|
|
.-
|
||
|
|
Authors: J. Scott Long and Jeremy Freese
|
||
|
|
www.indiana.edu/~jsl650/spost.htm
|
||
|
|
spostsup@@indiana.edu
|