calibrate_closeness

Finds the value of <code>closeness</code> (see <code>generate()</code>) for which synthetic data
reach a target value of a distance metric. The metric is evaluated on a
grid of closeness values with common random numbers, the grid interval
containing the target is located, and the solution is refined by
bisection, followed by a final linear interpolation step. Metric values
are themselves estimates, so the achieved value varies somewhat with the
random seed.
Available metrics (all computed against <code>data</code>, which should be the data
the synthesizer was fitted on):<ul>
<li><code>"auc"</code> -- cross-validated discriminator AUC (<code>discriminator_auc()</code>);
0.5 means indistinguishable, larger values mean more distant;</li>
<li><code>"dcr"</code> -- median distance to the closest real record relative to the
real-to-real nearest-neighbour distance (<code>dcr()</code>); larger means further
from real records;</li>
<li><code>"pmse"</code> -- pMSE ratio (<code>pmse()</code>) with the quadratic propensity model;</li>
<li><code>"marginal"</code> -- mean marginal distance (KS / TVD,
<code>marginal_metrics()</code>);</li>
<li>a function <code>function(real, synthetic)</code> returning a single number,
for any other criterion, e.g. a disclosure-risk measure from
<code>disclosure_risk()</code>.</li>
</ul>

Generates synthetic versions of numeric, categorical, mixed-type
and time-series data. Marginal distributions are modelled with
interpolated empirical quantile functions and dependence with Gaussian or
Student-t copulas (Sklar, 1959). Discrete and categorical variables enter
the copula through the distributional transform (Ruschendorf, 2009,
<doi:10.1016/j.jspi.2009.05.030>), and their latent correlations are
estimated by maximum likelihood as polychoric (Olsson, 1979,
<doi:10.1007/BF02296207>) and polyserial (Olsson, Drasgow and Dorans,
1982, <doi:10.1007/BF02294164>) correlations. Unordered categories are
placed on their latent coordinate in an order estimated from their
association with the other variables. Missing values are
modelled by missingness indicators inside the copula, so that
missingness may depend on other variables. Supports stratified models,
a closeness parameter with calibration to a target distance or
disclosure-risk level, chunked generation of very large data sets,
optional hybrid copula and conditional CART or GAM synthesis for
one continuous response (requiring 'rpart' or 'mgcv'), training-only
automatic selection of the conditional engine, experimental nominal
conditional synthesis of one unordered categorical response (requiring
'nnet'), copula vector-autoregressive and bootstrap time-series synthesis,
augmentation of small training sets for machine learning, and random
generation from multivariate distributions and elliptical and
Archimedean copulas. Fidelity, utility and disclosure risk are assessed
with marginal distances, association matrices, propensity score
mean-squared error (Snoke et al., 2018, <doi:10.1111/rssa.12358>),
confidence-interval overlap, correct attribution probability and
distance to closest record, with 'ggplot2' visualisations. Data can be
read from and written to CSV, TSV, RDS, RData, Parquet, Feather, HDF5,
JSON, NDJSON, Excel, SPSS, Stata, SAS and fst files. The
empirical-quantile and copula approach builds on ideas of the
'synthesizer' package by Mark van der Loo.

Bogdan Oancea

SynthesizerPlus

Synthetic Data Generation with Copulas, Evaluation and
Multi-Format I/O

calibrate_closeness function

<dl><dt>object</dt>
<dd>A fitted <code>sp_synthesizer</code>.</dd>
<dt>data</dt>
<dd>The real data the synthesizer was fitted on.</dd>
<dt>target</dt>
<dd>Target value of the metric.</dd>
<dt>metric</dt>
<dd>One of <code>"auc"</code>, <code>"dcr"</code>, <code>"pmse"</code>, <code>"marginal"</code>, or a function
<code>function(real, synthetic)</code> returning a number.</dd>
<dt>n</dt>
<dd>Number of synthetic records generated per evaluation (default:
size of <code>data</code>).</dd>
<dt>grid</dt>
<dd>Closeness values evaluated first.</dd>
<dt>refine</dt>
<dd>Number of bisection steps after the grid search.</dd>
<dt>max_rows</dt>
<dd>Maximum rows used by the metric computations.</dd>
<dt>seed</dt>
<dd>Random seed (the same seed is used at every closeness value).</dd>
<dt>...</dt>
<dd>Passed to <code>generate()</code> (e.g. <code>perturb</code> for directional
perturbations kept fixed during calibration).</dd></dl>

Arguments

Calibrate the closeness of synthetic to real data — calibrate_closeness

<dl>

<dt>object</dt>
<dd>A fitted <code>sp_synthesizer</code>.</dd>


<dt>data</dt>
<dd>The real data the synthesizer was fitted on.</dd>


<dt>target</dt>
<dd>Target value of the metric.</dd>


<dt>metric</dt>
<dd>One of <code>"auc"</code>, <code>"dcr"</code>, <code>"pmse"</code>, <code>"marginal"</code>, or a function
<code>function(real, synthetic)</code> returning a number.</dd>


<dt>n</dt>
<dd>Number of synthetic records generated per evaluation (default:
size of <code>data</code>).</dd>


<dt>grid</dt>
<dd>Closeness values evaluated first.</dd>


<dt>refine</dt>
<dd>Number of bisection steps after the grid search.</dd>


<dt>max_rows</dt>
<dd>Maximum rows used by the metric computations.</dd>


<dt>seed</dt>
<dd>Random seed (the same seed is used at every closeness value).</dd>


<dt>...</dt>
<dd>Passed to <code>generate()</code> (e.g. <code>perturb</code> for directional
perturbations kept fixed during calibration).</dd>

</dl>

calibrate_closeness: Calibrate the closeness of synthetic to real data

Description

Usage

Value

Arguments

Examples