{smcl}
{.-}
help for {cmd:oprobpr} {right:Nick Winter}
{.-}
{title:Displaying predicted probabilities from ordered probit & logit}
{p 5 15 2}
{cmdab:oprobpr}
{it:xvar}
[[{cmd:if} {it:exp}] [{cmd:in} {it:range}]
[{cmd:,}
{cmdab:l:evels(}{it:level_spec}{cmd:)}
{cmd:noesample}
{bind:{cmdab:cat:egories(}{it:cat_spec}[{cmd:,} ...]}{cmd:)}
{bind:{cmdab:lab:els(}{it:lablist}{cmd:)}}
{bind:{cmdab:gt:ype(}{it:plot_type}{cmd:)}}
{cmdab:stack}
{cmdab:f:rom(}{it:#}{cmd:)} {cmdab:t:o(}{it:#}{cmd:)} {cmdab:inc(}{it:#}{cmd:)}
{bind:{cmdab:nwid:th(}{it:#}{cmd:)}}
{cmdab:nol:ist}
{bind:{cmdab:save:(}{it:filename}[{cmd:, replace}]}{cmd:)}}
{bind:{cmd:plot(}{it:plot}{cmd:)}}
{it:graph_options}
{title:Description}
{p 0 5 2}
{cmd:oprobpr} plots (and optionally lists) the predicted probabilities from a previously-estimated
ordered dependent variable model against a single covariate from the model,
holding the other covariates constant.
{p 0 5 2}
By default all response categories are listed and plotted; the {cmd:categories()}
option allows the user to control plotting of only some categories, or
combinations of categories. {cmd:labels()} allows control of the labelling
of the plotted lines.
{p 0 5 2}
{cmd:oprobpr} is a substantially-updated version that takes advantage of Stata's version
8 graphics. With this revision, {cmd:oprobpr} now operates as a post-estimation command,
rather than estimating and plotting the model in one step. See the important note below on handling
interaction terms in the revised command. This program was originally based
on {help logpred} published by Joanne Garrett
and of {help probpred} published by Mead Over in sg42.2: STB 42.
{title:Options}
{p 0 5 2}
{cmd:levels(}{it:level_spec}{cmd:)} specifies the levels at which to hold the other covariates
in the model. By default, all covariates are set at their estimation sample means (but see the {cmd:noesample()} option).
Note that the command automatically calculates these means in the estimation sample only , any {cmd:if} and/or {cmd:in} conditions
specified with {cmd:oprobpr} will {it:further} restrict the calculation of these means.
{p 5 5 2}The {cmd:levels()} option allows you to set covariates to other values; e.g. {cmd:levels(mpg=50, foreign=1)}.
{p 5 5 2}In addition, if the model includes interaction terms between the {it:xvar} variable you are plotting and some other
covariate, they must be appropriately specified
in the {cmd:levels()} option in order that {cmd:oprobpr} can re-calculate them appropriately. For example, if
your model included a variable called mXw, which is the interaction between mpg and weight,
and mpg is your {it:xvar}, then you would specify {cmd:levels(mXw=mpgXweight)}. Of course, you can mix
and match the two uses of {cmd:levels()}: {cmd:levels(weight=2500, mXw=mpg*weight)}.
{p 0 5 2}
{cmd:noesample} specifies that the calculation of means of covariates should be done without
regard to the estimation sample from the model estimation. This means both that any {cmd:if} and/or {cmd:in}
conditions specified when the model was estimated are ignorred, and that all cases are included in the
calculation of each variable's mean (that is, cases are not excluded on a casewise basis). Any {cmd:if} and/or {cmd:in}
conditions specified with the {cmd:oprobpr} do restrict the calculation of means.
{p 0 5 2}
{cmd:from(}{it:#}{cmd:)} specifies the lowest value of {it:xvar} for which a prediction is to be
calculated. The default is to use the minimum of {it:xvar} in the estimation sample.
{p 0 5 2}
{cmd:to(}{it:#}{cmd:)} specifies the highest value of {it:xvar} for which a prediction is to be
calculated. The default is to use the maximum of {it:xvar} in the estimation sample.
{p 0 5 2}
{cmd:newobs(}{it:#}{cmd:)} specifies the number of observations to be created for calculation and graphing
of predicted probabilities. The default is 25; more may be specified if necessary to yield a smooth line.
{p 0 5 2}
{cmd:categories(}{it:n,...,n}{cmd:)} controls which categories of the dependent variable
are plotted and listed. The default is to list and plot probabilities
for all categories. For example, {cmd:cat(1,3,4)} would result in categories
1, 3 and 4 only being listed and plotted.
{p 5 5 2}
{cmd:categories()} also allows categories to be combined. So, for example,
{cmd:cat(1+2,3,4+5+6)} would plot three lines: one that is the sum of probabilities
for categories one and two, one that is the probability of category three,
and one that is the sum of categories 4 through 6.
{p 0 5 2}{cmd:labels()} specifies text labels with which to label the lines.
By default, simple categories are labeled with the appropriate
value label from the dependent variable, if available. Otherwise, they are
labelled {bind:"Category 1"}, {bind:"Category 2"} ... , through {bind:"Category {it:n}"}.
For example,
{cmd:cat(Low Medium High)} would label the lines "Low", "Medium",
and "High". To leave a line unlabeled, indicate a "." for its label.
{p 5 5 2}Enclose any labels that have spaces with quotation marks: {cmd:lab("Very Low" Low Medium High "Very High")}
{p 5 5 2}Note that the labels can also be specified within the {help legend_option:legend()} option's {cmd:label()} suboption;
this just provides a somewhat easier way to change multiple labels with a single option.
{p 0 5 2}
{cmd:gtype()} specifies the type of twoway graph to create. By default this is {help scatter:scatter}.
{p 0 5 2}
{cmd:stack} causes the categories to be calculated as cumulative probabilities, allowing the
creation of a stacked plot; when this is specified the {cmd:gtype()} defaults to {help twoway_area:area}.
{p 0 5 2}
{cmd:nwidth()} specifies the width of the note that {cmd:oprobpr} creates to indicate the levels of the
covariates. The default is 80.
{p 0 5 2}
{cmd:save(}{it:filename}{cmd:)} saves the prediction data set. This is useful
for conducting additional analysis of the predicted values. (Note that the
{help graph} option {cmd:saving()} is different, and may be used to save
the resulting {it:graph}.)
{p 0 5 2}
{cmd:plot(}{it:plot}{cmd:)} provides a way to add other plots to the generated
graph; see help {help plot_option}. Note that the dataset will be intact, but will have
additional observations added at the end for the calculation of predicted probabilities. To exclude
these observations from any additional plots, include the condition {cmd:in @@@} in these plots; this will
be translated to cover the range of the original dataset.
{p 0 5 2}
{it:graph_options} can be any valid options for a twoway graph.
{title:Examples}
{p 5 10 2}
{cmd:. oprobit rep78 mpg weight gear_ratio foreign}
{cmd:. oprobpr mpg}
{p 5 5 2}
Plots the predicted probabilities for categories of rep78 against mpg, holding the
other covariates constant at their estimation-sample means.
{p 5 10 2}
{cmd:. oprobpr mpg, levels(weight=2500, foreign=0)}
{p 5 5 2}
Same as above, except predictions are for foreign==0 and weight==2500 instead of for
the sample averages of those variables.
{p 5 10 2}
{cmd:. generate mpgXweight = mpg*weight}
{p_end}{p 5 10 2}
{cmd:. ologit rep78 mpg weight mpgXweight gear_ratio foreign}
{p_end}{p 5 10 2}
{cmd:. oprobpr mpg, levels(weight=2500, mpgXweight=mpg*weight, foreign=0)}
{p 5 5 2}
Same as above, except that the interaction term weight*mpg is included
in the original model, and is specified in the {cmd:oprobpr()} command. Weight is held
constant at 2500, foreign at 0, and gear_ratio at its sample mean. Model estimated is ordered
logit rather than ordered probit.
{p 5 10 2}
{cmd:. oprobpr mpg, cat(1,3+4,5) lab(Low "Medium & High" "Very High") clpattern(dash dot dash_dot) msym(i i i) title( , size(small)) }
{p 5 5 2}
Plots only categories 1, the sum of
3 and 4, and 5 of rep78, and labels them "Low", "Medium & High", and "Very High",
respectively. The default {help connect_options:line patterns} and {help scatter:marker symbols}
are overridden, and the size of the overall {help title_options:title} is adjusted.
Note that any other option
which works on the {help graph} command will also work here.
{p 5 10 2}
{cmd:. oprobpr mpg, stack }
{p 5 5 2}
Probabilities are graphed as a stacked area plot instead of as a series of lines.
{p 5 10 2}
{cmd:. oprobpr mpg, plot(function y=.75, range(mpg) ) legend(order(1 2 3 4 5)}
{p 5 5 2}
An additional plot is overlaid, in this case a horizontal line at 0.75, specified to run over the
range of the variable mpg. The legend option is specified to suppress the listing of this additional
plot in the legend. Note that this effect could also be achieved with the {cmd:yline()} option.
{title:Author}
Nicholas Winter
University of Virginia
nwinter@virginia.edu
{title:Also see}
{p 0 10 2}
On-line: help for {help predict}, {help oprobit}, {help ologit},
{help graph twoway}
{p_end}