pastas.stats.metrics.rsq
========================

.. py:function:: pastas.stats.metrics.rsq(obs: pandas.Series, sim: pandas.Series | None = None, res: pandas.Series | None = None, missing: str = 'drop', weighted: bool = False, max_gap: int = 30, nparam: int | None = None) -> float

   
   Compute R-squared, possibly adjusted for the number of free parameters.

   :param obs: Series with the observed values.
   :type obs: pandas.Series
   :param sim: The Series with the simulated values.
   :type sim: pandas.Series, optional
   :param res: The Series with the residual values. If time series for the residuals are
               provided, the sim and obs arguments are ignored. Note that the residuals
               must be computed as `obs - sim` here.
   :type res: pandas.Series, optional
   :param missing: string with the rule to deal with missing values. Only "drop" is supported now.
   :type missing: str, optional
   :param weighted: If weighted is True, the variances are computed using the time step between
                    observations as weights. Default is False.
   :type weighted: bool, optional
   :param max_gap: maximum allowed gap period in days to use for the computation of the weights.
                   All time steps larger than max_gap are replace with the max_gap value.
                   Default value is 30 days.
   :type max_gap: int, optional
   :param nparam: number of calibrated parameters.
   :type nparam: int, optional

   .. rubric:: Notes

   .. math:: \rho_{adj} = 1-  \frac{n-1}{n-n_{param}}*\frac{rss}{tss}

   Where n is the number of observations, :math:`n_{param}` the number of free
   parameters, rss the sum of the squared errors, and tss the total sum of
   squared errors.

   When nparam is provided, the :math:`\rho` is adjusted for the number of
   calibration parameters.















   ..
       !! processed by numpydoc !!