process_qs

Processes the inputs to <code>optimize_controls()</code> to formulate sample size constraints.

internal

Natural strata can be used in observational studies to balance
the distributions of many covariates across any number of treatment
groups and any number of comparisons. These strata have proportional
amounts of units within each stratum across the treatments, allowing
for simple interpretation and aggregation across strata. Within each
stratum, the units are chosen using randomized rounding of a linear
program that balances many covariates. For more details, see Brumberg et al. (2022) <doi:10.1111/rssa.12848> and Brumberg et al.(2023) <doi:10.1093/jrsssc/qlad010>.
To solve the linear program, the 'Gurobi' commercial optimization software
is recommended, but not required. The 'gurobi' R package can be installed by following the instructions at <https://docs.gurobi.com/projects/optimizer/en/current/reference/r/setup.html> after claiming your free academic license at <https://www.gurobi.com/academia/academic-program-and-licenses/>.

Katherine Brumberg

natstrat

Obtain Unweighted Natural Strata that Balance Many Covariates

process_qs function

<dl><dt>ratio</dt>
<dd>a numeric or vector specifying the desired ratio of controls to `treated` in
each stratum. If there is one control group and all treated units should be included,
this can be a numeric. Otherwise, this should be
a vector with one entry per treatment group, in the same order as the levels of
<code>z</code>, including the treated level. If <code>NULL</code>, <code>q_s</code> should be specified.</dd>
<dt>q_s</dt>
<dd>a named vector or matrix indicating how many units are to be selected from each stratum.
If there is one control group and all treated units are desired, this can be a vector; otherwise,
this should have one row per treatment group, where the order of the rows matches the order of
the levels of <code>z</code>, including the treated level.
If <code>NULL</code>, <code>ratio</code> should be specified. If both are specified, <code>q_s</code> will take priority.
Typically, if the desired ratio is not feasible for every stratum, <code>q_s</code> should be generated
using <code>generate_qs()</code>.</dd>
<dt>treated</dt>
<dd>which treatment value should be considered the treated units. This
must be one of the values of <code>z</code>.</dd></dl>

Arguments

Process the desired sample sizes for optimize_controls() — process_qs

<dl>

<dt>ratio</dt>
<dd>a numeric or vector specifying the desired ratio of controls to `treated` in
each stratum. If there is one control group and all treated units should be included,
this can be a numeric. Otherwise, this should be
a vector with one entry per treatment group, in the same order as the levels of
<code>z</code>, including the treated level. If <code>NULL</code>, <code>q_s</code> should be specified.</dd>


<dt>q_s</dt>
<dd>a named vector or matrix indicating how many units are to be selected from each stratum.
If there is one control group and all treated units are desired, this can be a vector; otherwise,
this should have one row per treatment group, where the order of the rows matches the order of
the levels of <code>z</code>, including the treated level.
If <code>NULL</code>, <code>ratio</code> should be specified. If both are specified, <code>q_s</code> will take priority.
Typically, if the desired ratio is not feasible for every stratum, <code>q_s</code> should be generated
using <code>generate_qs()</code>.</dd>


<dt>treated</dt>
<dd>which treatment value should be considered the treated units. This
must be one of the values of <code>z</code>.</dd>

</dl>

process_qs: Process the desired sample sizes for `optimize_controls()`

Description

Usage

Value

Arguments