##################################################################
Extending an inversion
##################################################################

Two cases are covered here:

  - an inversion that stopped unexpectedly, e.g. due to a timeout, a full disk, or another such failure
  - an inversion that stopped normally but needs to be extended, e.g. because the maximum number of simulations was reached or tighter convergence is required


1) If the inversion stopped unexpectedly, remove the last directory generated by the CTM (e.g. for CHIMERE with 24-hour periods, remove the last day's directory).
2) Modify the YAML to reuse the valid computations:

   - in :bash:`obsoperator`, set :bash:`autorestart` to ``True``
   - in :bash:`simulator`, set :bash:`reload_from_previous` to ``True``. If :bash:`simulator` is not specified in the YAML (which is the usual case when using the default cost function), add it as shown in the example below.
   
.. container:: toggle

  .. code-block:: yaml
  
      mode:
        minimizer:
          df1: 0.5
          epsg: 0.1
          maxiter: 10
          nsim: 25
          plugin:
            name: M1QN3
            version: std
          simulator:
            plugin:
              name: gausscost
              version: std
            reload_from_previous: true
        plugin:
          name: 4dvar
          version: std
        save_out_netcdf: true
      obsoperator:
        plugin:
          name: standard
          version: std
          type: obsoperator
        autorestart: True

3) To extend the inversion, adjust the minimizer's stopping criteria. For M1QN3, decrease :bash:`epsg` and increase :bash:`maxiter` and :bash:`nsim`.
4) Relaunch the job.

.. ##### lien avec ceci: pour relancer une inv en cours: supprimer le répertoire dans lequel le modèle s'est arreté (ex: adj_0015) et relancer avec autorestart dans obsoper et reload_results dans mode