dmean

For each numerical variable in the two datasets, compute the relative
difference between the mean (standard deviation) of the real data and the
mean (standard deviation) of the synthetic data. The summary is the average
of these relative differences over all numerical variables.

Data is synthesized using a combination of inverse transform
sampling from the empirical quantile functions for each variable, and
then copying the rank order structure from the original dataset. The
package also includes a number of functions to measure the utility of
synthesized datasets.

Mark der Loo

synthesizer

Synthesize Data Based on Empirical Quantile Functions and Rank
Order Matching

Mark van der Loo

dmean function

<dl><dt>synth</dt>
<dd><code>[data.frame]</code> Synthetic data</dd>
<dt>real</dt>
<dd><code>[data.frame]</code> Real data</dd>
<dt>tol</dt>
<dd><code>[numeric]</code> Nonnegative tolerance. If the absolute 
mean (standard deviation) of a variable is smaller than 
<code>tol</code>, it is considered zero. In that case the 
absolute difference instead of the absolute relative 
difference is computed.</dd>
<dt>...</dt>
<dd>Arguments passed to <code>mean</code>. e.g. use <code>trim=c(0.01,0.99)</code>
for mean estimation that is less sensitive to outliers.</dd></dl>

Arguments

Summarization of location and spread between synthetic and real data — dmean

<dl>

<dt>synth</dt>
<dd><code>[data.frame]</code> Synthetic data</dd>


<dt>real</dt>
<dd><code>[data.frame]</code> Real data</dd>


<dt>tol</dt>
<dd><code>[numeric]</code> Nonnegative tolerance. If the absolute 
mean (standard deviation) of a variable is smaller than 
<code>tol</code>, it is considered zero. In that case the 
absolute difference instead of the absolute relative 
difference is computed.</dd>


<dt>...</dt>
<dd>Arguments passed to <code>mean</code>. e.g. use <code>trim=c(0.01,0.99)</code>
for mean estimation that is less sensitive to outliers.</dd>

</dl>

dmean: Summarization of location and spread between synthetic and real data

Description

Usage

Value

Arguments

See Also

Examples