.update_obs

<code>.update_obs</code> is an experimental utility to recycle the information in an inference object to produce an inference for different data without repeating a simulation workflow. Beyond its experimental nature, its result is not expected to provide the same precision of inference as a standard iterative Infusion workflow, since the recycled simulations were adapted to the original data only (<code>.update_obs</code> indeed allows one to investigate the effect of using this non-adapted information). So the new results should not be used as a equivalent to a full Infusion iterative workflow.

Implements functions for simulation-based inference. In particular, implements functions to perform likelihood inference from data summaries whose distributions are simulated. The package implements more advanced methods than the ones first described in: Rousset, Gouy, Almoyna and Courtiol (2017) <doi:10.1111/1755-0998.12627>.

Franois Rousset

Infusion

Inference Using Simulation

François Rousset

.update_obs function

<dl> <dt>object</dt>
<dd>an object of class <code>"SLik_j"</code> as createded by <code>infer_SLik_joint</code> (possible updated by <code>MSL</code> and <code>refine</code>).</dd> <dt>new.obs</dt>
<dd>A named vector of summary statistics (ideally the projected ones, but raw ones are handled)</dd> <dt>CIs, eval_RMSEs, ...</dt>
<dd>further arguments passed to <code>MSL</code>.</dd></dl>

Arguments

Updating an 'SLik_j' object for new data — .update_obs

<dl>

 <dt>object</dt>
<dd>an object of class <code>"SLik_j"</code> as createded by <code>infer_SLik_joint</code> (possible updated by <code>MSL</code> and <code>refine</code>).</dd>

 <dt>new.obs</dt>
<dd>A named vector of summary statistics (ideally the projected ones, but raw ones are handled)</dd>

 <dt>CIs, eval_RMSEs, ...</dt>
<dd>further arguments passed to <code>MSL</code>.</dd>

</dl>

.update_obs: Updating an 'SLik_j' object for new data

Description

Usage

Value

Arguments