pycif.plugins.datastreams.meteos.chimere_meteo — API reference

Configuration reference: chimere_meteo plugin

pycif.plugins.datastreams.meteos.chimere_meteo.fetch.fetch(ref_dir, ref_file, input_interval, target_dir, tracer=None, **kwargs)

Locate CHIMERE METEO.nc files and link them to the working directory.

Iterates over period-start dates covering input_interval at tracer.file_freq, formats ref_dir/ref_file for each period, and for each file that exists on disk records the hourly sub-periods it covers and symlinks it into target_dir. Periods whose file does not exist are silently skipped (absent from the returned dicts).

Parameters:

• ref_dir – directory holding the CHIMERE METEO.nc files.

• ref_file – strftime-style file name pattern (relative to ref_dir), formatted once per period.

• input_interval – 2-element sequence (datei, datef) giving the requested date range.

• target_dir – directory where matching files are symlinked.

• tracer – the meteo datastream Plugin, used for tracer.file_freq.

Returns:

(list_files, list_dates), dicts keyed by each period’s start date. list_files maps each key to the (repeated) linked file path, one entry per hour in the period; list_dates maps each key to the corresponding list of [hour, hour + 1h] date pairs.

Return type:

tuple

pycif.plugins.datastreams.meteos.chimere_meteo.read.read(self, name, varnames, dates, files, interpol_flx=False, comp_type=None, ddi=None, **kwargs)

Read requested variables from CHIMERE METEO.nc files.

For each [period, file] pair, reads varnames (falling back to name if varnames is empty) and the Times variable from the file, parses Times to locate the array index matching the requested period start (optionally offset from ddi instead of the file’s own first time stamp), and extracts that hourly slice.

Parameters:

• self – the meteo Plugin (uses self.model.periods when raising the interpolation-misconfiguration error).

• name – name of the component/variable to read, used as a fallback when varnames is empty.

• varnames – name of the NetCDF variable to extract; must not be an empty string.

• dates – list of [start, end] date pairs, aligned with files.

• files – list of CHIMERE METEO.nc files to read from, aligned with dates.

• interpol_flx (bool) – unused here (kept for interface consistency).

• comp_type – unused here (kept for interface consistency).

• ddi – optional reference date used instead of the file’s first Times entry to compute the hourly index to extract.

Returns:

extracted data stacked over dates, with dims (time, lev, lat, lon).

Return type:

xarray.DataArray

Raises:

CifValueError – if varnames is an empty string, which happens when the meteo needs interpolating to another domain or temporal resolution but the input datavect does not list all required variables explicitly.

pycif.plugins.datastreams.meteos.chimere_meteo.write.write(self, name, meteo_file, meteo_data, mode='a', **kwargs)

Write meteo/flux data to a CHIMERE-compatible METEO.nc file.

Switches to write mode if meteo_file does not already exist, drops a singleton lev dimension, renames dimensions to CHIMERE conventions (Time, bottom_top, south_north, west_east), builds the CHIMERE Times character-string variable, injects the domain’s lon/lat and sigma vertical coordinates, duplicates the time axis when writing nphourm (to accelerate its reading), and, when self.ncformat is "NETCDF4", applies zlib/float32 compression and chunks the data before dumping to disk.

Parameters:

• self (Fluxes) – the meteo datastream Plugin (uses self.domain for coordinates and self.ncformat for the output format).

• name (str) – name of the variable/component being written; special handling applies when equal to "nphourm".

• meteo_file (str) – the file where to write the meteo/flux data.

• meteo_data (xarray.Dataset) – meteo/flux data to write.

• mode (str) – "w" to overwrite, "a" to append; forced to "w" when meteo_file does not already exist.