(calibration_markov_chain_results_parameters)=
# Drawing the parameters

The posterior distributions of the parameters can be visualised using the dedicated instance of 
`drawParameters`, whose prototype is identical to that discussed in 
[](#calibration_classes_functions_observations_calibration_classes_draw_parameter).

````{only} cpp
```cpp
void drawParameters(TString sTitre, const char *variable = "*", Option_t * option = "");
```
````

````{only} py
```python
void drawParameters(sTitre, variable = "*", option = "")
```
````

The only difference from the standard `drawParameters` method in `TCalibration` is that this 
version automatically accounts for the burn-in period and the lag (see
[](#calibration_markov_chain_diagnostics) for details). Specifically, the first samples of each 
chain are discarded according to the burn-in value, and the remaining samples are thinned using 
the specified lag (i.e., one sample is retained out of every lag iterations).

An example of this function is shown in {numref}`usecases_calibrationMCMCLinReg_Par`.