…eparately from each obs.field entry because the obs dictionary does not exist in the GenEnsProd config file.

…can perform lookups for the parent.

…s that retrieve the climatology data. Rather than requiring all the climo_mean and climo_stdev dictionary entries to be defined at the same config file context level, parse each one individually. This enables the METplus wrappers to only partially override this dictionary and still rely on the default values provided in MET's default configuration files.

…d on the new function signature. Also update the tools to check the number of climo fields separately for the forecast and observation climos.

…_config

…nary. Use config.fcst.climo_mean.regrid first, config.fcst.regrid second, and config.climo_mean.regrid third. Notably, DO NOT use config.regrid. This is definitely the problem with having regrid specified at mutliple config file context levels. It makes the logic for which to use when very messy.

… the climatology regridding logic inside fcst is problematic because it applies to the forecast data as well and you end up with the verification grid being undefined. So the climo regridding logic must be defined in 'climo_mean.regrid' either within the 'fcst' and 'obs' dictionaries or at the top-level config context.

…_config

…t, Upper_Right, Lower_Right, and Lower_Left interpolation methods to the list of valid options for regridding, as already indicated in the MET User's Guide.

…_config

…e it work the way @georgemccabe expects it to. It now uses pointers to both the primary and default dictionaries and parses each entry individually.

… the regrid dictionary.

…_config

…ing' log message to avoid confusion.

…_config

…literals.

…umber of climo fields provided. If there's 0, just return since no data has been requested. If there's 1, use it regardless of the number of input fields. If there's more than 1, just use the requested i_vx index value.

…ooping over the series entries.

…strings for consistency.

…is PR but works after the changes in this PR.