{smcl}
{hline}
help for {cmd:factref} {right:(Roger Newson)}
{hline}
{title:Insert new observations with reference values for factors}
{p 8 15}{cmd:factref} {it:varlist} [{cmd:if} {it:exp}] [{cmd:in} {it:range}] [ {cmd:,}
{cmd:by(}{it:varlist}{cmd:)} {cmdab:re:find}{cmd:(}{it:newvarname}{cmd:)}
{cmdab:rz:ero}{cmd:(}{it:varlist}{cmd:)} {cmdab:ru:nity}{cmd:(}{it:varlist}{cmd:)}
]
{title:Description}
{p}
{cmd:factref} is used with {help factext}, {help descsave}, {help parmby} and {help parmest}.
It takes, as input, a data set created by {help parmby} or {help parmest}, in which factors
have been created by {help descsave} and {help factext}. These factors are variables (numeric or string)
representing categorical variables in the fitted model, represented by dummy variables such as those
created by {help xi}. {cmd:factref} adds to that data set a set of additional observations
with reference values for these factors, copied from the {cmd:omit} characteristic for each factor.
(The {cmd:omit} characteristic is used by the {help xi} command to store the reference value
of a categorical variable; see {hi:[R] xi}.)
For each factor, {cmd:factref} creates one reference observation if there are non-missing values for the factor
and there is no by-variable list, or one reference observation for each by-group where there are non-missing
values of the factor if there is a by-variable list.
After the reference observations have been added, the user may create a plot or table of confidence intervals
for factor effects.
{title:Options}
{p 0 4}{cmd:by(}{it:varlist}{cmd:)} specifies a list of by-variables. {cmd:factref} will create a reference
observation for each factor for each by-group containing non-missing values of that factor.
{p 0 4}{cmd:refind(}{it:newvarname}{cmd:)} specifies an output variable indicating, in each observation,
whether that observation is a reference observation. It contains a missing value if that observation
is excluded by the {cmd:if} and/or {cmd:in} statements, a value of 1 if that observation is a reference observation,
and a value of 0 otherwise.
{p 0 4}{cmd:rzero(}{it:varlist}{cmd:)} specifies a list of existing variables whose values in the reference
observations are to be set to 0 (if the variable is numeric) or "0" (if the variable is string).
These variables might represent estimates and confidence limits for difference parameters, eg additive
factor effects such as mean differences between factor levels.
{p 0 4}{cmd:runity(}{it:varlist}{cmd:)} specifies a list of existing variables whose values in the reference
observations are to be set to 1 (if the variable is numeric) or "1" (if the variable is string).
These variables might represent estimates and confidence limits for ratio parameters, eg multiplicative
factor effects such as relative risks between factor levels.
{title:Remarks}
{p}
{cmd:factref} is typically used with {help factext}, which is used with {help parmby}, {help parmest} and
{help descsave} to create a data set with one observation per model parameter and data on confidence intervals
and P-values. {cmd:factref} adds, to such a data set, a set of reference observations, one for each factor,
or one for each by-group containing at least one non-missing value of each factor if the {cmd:by()} option
is specified. A reference observation for a factor has a value for that factor taken from the {hi:omit}
characteristic for that factor. The other variables in a reference observation may have values reset
to zero or one if the variable is listed in the options {cmd:rzero} or {cmd:runity},
respectively. Otherwise, the value of any variable in a reference observation, apart from the factors and
the by-group variables specified in the {cmd:by()} option, is set to missing.
For {cmd:factref} to work, the {cmd:omit} characteristic for each factor must be set to the reference value
for that factor. This can be done by specifying the {cmd:omit} characteristic for the factor of the same name
in the original data set, and then using the {cmd:char(omit)} option with {help descsave}.
(The {cmd:omit} characteristic is used by the {help xi} command; see {hi:[R] xi}.)
{title:Examples}
{p}
The following example will work with the {hi:auto} data
if {help parmest}, {help parmby}, {help descsave} and {help factext} are installed.
A logistic regression is carried out, comparing the odds of non-US origin between
3 tertiles of mileage, created by the {help xtile} command. The odds ratios (with confidence limits)
are plotted against mileage tertile. Note that the characteristic {cmd:mpggp[omit]}, used optionally by {help xi},
must be specified if {cmd:factref} is to work.
{p 16 20}{inp:. xtile mpggp=mpg, nq(3)}{p_end}
{p 16 20}{inp:. char mpggp[omit] 1}{p_end}
{p 16 20}{inp:. tempfile tf0}{p_end}
{p 16 20}{inp:. descsave ,char(omit) do(`"`tf0'"', replace)}{p_end}
{p 16 20}{inp:. parmby "xi: logit foreign i.mpggp,or", label norestore eform format(estimate min* max* %8.2f p %8.1e)}{p_end}
{p 16 20}{inp:. factext, do(`"`tf0'"')}{p_end}
{p 16 20}{inp:. factref mpggp, ru(estimate min* max*)}{p_end}
{p 16 20}{inp:. list}{p_end}
{p 16 20}{inp:. version 7: graph estimate min95 max95 mpggp, s(O..) c(.II) xlab(1 2 3) ylab ylin(1)}{p_end}
{p}
The following example illustrates the {cmd:by()} option, and will work with the {hi:auto} data
if {help parmest}, {help parmby}, {help descsave} and {help factext} are installed.
It will add reference observations to the {help parmby} output
data set and create a graph of confidence intervals. Note that the characteristic
{cmd:rep78[omit]}, used optionally by {help xi}, must be specified if {cmd:factref} is to work.
{p 16 20}{inp:. char rep78[omit] 1}{p_end}
{p 16 20}{inp:. tempfile tf0}{p_end}
{p 16 20}{inp:. descsave, char(omit) do(`"`tf0'"', replace)}{p_end}
{p 16 20}{inp:. parmby "xi:regress mpg i.rep78", label by(foreign) norestore format(estimate min* max* %8.2f p %8.1e)}{p_end}
{p 16 20}{inp:. factext, do(`"`tf0'"')}{p_end}
{p 16 20}{inp:. factref rep78, by(foreign) rz(estimate stderr min* max*) refind(refcat)}{p_end}
{p 16 20}{inp:. list}{p_end}
{p 16 20}{inp:. version 7: graph estimate min95 max95 rep78, s(O..) c(.II) by(foreign) xlab ylab ylin(0)}{p_end}
Another example of the use of {cmd:factref} can be found in Newson (2003).
{title:Author}
{p}
Roger Newson, Imperial College London, UK.
Email: {browse "mailto:r.newson@imperial.ac.uk":r.newson@imperial.ac.uk}
{title:References}
{p}Newson, R. 2003. Confidence intervals and {it:p}-values for delivery to the end user.
{it:The Stata Journal} 3(3): 245-269. A pre-publication draft is downloadable from
{net "from http://www.imperial.ac.uk/nhli/r.newson/":Roger Newson's website at http://www.imperial.ac.uk/nhli/r.newson/}
{title:Also see}
{p 0 10}
{bind: }Manual: {hi:[R] xi}, {hi:[R] pctile}.
{p_end}
{p 0 10}
On-line: help for {help describe}, {help tabulate}, {help xi}, {help graph}
{break} help for {help parmby}, {help parmest}, {help descsave}, {help factext} and {help factmerg} if installed
{p_end}