slv.inversion.report.plot_obs#

slv.inversion.report.plot_obs(post, location=None, location_level='obs_location', rolling_window='30D', gap_threshold='7D', show_raw=True)[source]#

Plot the obs against what the prior and the posterior model, one figure per location.

Three series over time: the observed concentrations, the posterior’s modeled ones (post.modeled_obs) and the prior’s (problem.predict(problem.prior)), each as points and as a rolling mean. A gap in a location’s times longer than gap_threshold is shaded, and the rolling means break there.

Parameters:

  • post (Posterior) – The solved inversion.

  • location (label or list of labels, optional) – The location (or locations) to plot. By default, every location.

  • location_level (str, default 'obs_location') – The level of the obs’ labels that holds the locations.

  • rolling_window (str or int, default '30D') – The window of the rolling means, as DataFrame.rolling takes it.

  • gap_threshold (str, default '7D') – The shortest gap between two times that is shaded as missing data.

  • show_raw (bool, default True) – Whether to draw the points behind the rolling means.

Returns:

The axes, one per location.

Return type:

numpy.ndarray