GenerateSyntheticData

Generating the synthetic RCT data given marginal distribution of each covariate

Validates estimates of (conditional) average treatment effects obtained using observational data by a) making it easy to obtain and visualize estimates derived using a large variety of methods (G-computation, inverse propensity score weighting, etc.), and b) ensuring that estimates are easily compared to a gold standard (i.e., estimates derived from randomized controlled trials). 'RCTrep' offers a generic protocol for treatment effect validation based on four simple steps, namely, set-selection, estimation, diagnosis, and validation. 'RCTrep' provides a simple dashboard to review the obtained results. The validation approach is introduced by Shen, L., Geleijnse, G. and Kaptein, M. (2023) <doi:10.21203/rs.3.rs-2559287/v2>.

Lingjie Shen

RCTrep

Validation of Estimates of Treatment Effects in Observational
Data

Gijs Geleijnse

Maurits Kaptein

GenerateSyntheticData function

<dl><dt>margin_dis</dt>
<dd>a character indicating the distribution of each variable, allowable options are <code>"bernoulli_categorical"</code> and <code>"bernoulli"</code>. If some variables have two categories and some have more than two categories, <code>"bernoulli_categorical"</code> should be specified; if all variables have two categories, <code>"bernoulli"</code> should be specified</dd>
<dt>N</dt>
<dd>a numeric value specifying the sample size for the simulated data</dd>
<dt>margin</dt>
<dd>a list containing the marginal distribution of variables; if margin_dis="bernoulli_categorical", then margin should be list(x1=c("x1",nlevels(x1),level1, level2,...,leveln, plevel1, plevel2,...,plevel3), x2=c("x2",...)); if margin_dis="bernoulli", margin=list(p(x1=1),p(x2=1),...,p(xn=1))</dd>
<dt>var_name</dt>
<dd>a vector indicating the name of variables, the order of variables should be aligned with <code>margin</code></dd>
<dt>pw.cor</dt>
<dd>a vector specifying the pairwise correlations of the variables, default is 0; when margin_dis="bernoulli", then pw.cor must be specified.</dd></dl>

Arguments

Generating the synthetic RCT data given marginal distribution of each covariate — GenerateSyntheticData

<dl>

<dt>margin_dis</dt>
<dd>a character indicating the distribution of each variable, allowable options are <code>"bernoulli_categorical"</code> and <code>"bernoulli"</code>. If some variables have two categories and some have more than two categories, <code>"bernoulli_categorical"</code> should be specified; if all variables have two categories, <code>"bernoulli"</code> should be specified</dd>


<dt>N</dt>
<dd>a numeric value specifying the sample size for the simulated data</dd>


<dt>margin</dt>
<dd>a list containing the marginal distribution of variables; if margin_dis="bernoulli_categorical", then margin should be list(x1=c("x1",nlevels(x1),level1, level2,...,leveln, plevel1, plevel2,...,plevel3), x2=c("x2",...)); if margin_dis="bernoulli", margin=list(p(x1=1),p(x2=1),...,p(xn=1))</dd>


<dt>var_name</dt>
<dd>a vector indicating the name of variables, the order of variables should be aligned with <code>margin</code></dd>


<dt>pw.cor</dt>
<dd>a vector specifying the pairwise correlations of the variables, default is 0; when margin_dis="bernoulli", then pw.cor must be specified.</dd>

</dl>

GenerateSyntheticData: Generating the synthetic RCT data given marginal distribution of each covariate

Description

Usage

Value

Arguments