decimal_places: from ``mdf_reader`` to ``cdm_mapper ``

[ludwiglierhammer]

Move information about decimal_places from mdf_reader to cdm_mapper as they are only used in cdm_mapper.

[ludwiglierhammer]

Hi @aanderss, I found some inconsistencies in the decimal places. Could you please takea look at the list below?

701/721

The precision of original_value in the output data is not equal to the precision in the input data. There are five decimal places in the output and none in the input. original_value and observation_value are not similar.

e.g. (721):

• (c99_data, air_temperature) is 53 -> original_value is 53.00000
• core, AT is 44 -> observation_value is 277.55

✔️

702

In the observations tables there are no original_values defined. Should we set them to observation_value like here?

714

In the observations-wd table decimal_places of original_value is set to 0. Should we change the value to 2 like in the input data?

✔️

730

empty observations_tabless: AT, SST, DPT and SLP. Given decimal places: 5
Should we set this to 2?

✔️ -> according to backup files on rcl.

781

In the observations-slp table decimal_places of original_value is set to 2. Should we change the value to 1 like in the input data (same with observations-ws).

✔️

C-RAID

In the header table decimal_places of height_of_station_above_local_ground, height_of_station_above_sea_level and report_time_accuracy are not set. If decimal_places are not set, they will be automatically set to 5. Should we set them equal to here(1, 1, 0)

✔️

[ludwiglierhammer]

702

In the observations tables there are no original_values defined. Should we set them to observation_value like here?

original_units is taken from c99_data attachment.
This attachment does NOT include any elements for original_value. Thus, original_value is not defined in den cdm_mapper.

@aanderss: Should we leave it as it is or take core elements for both original_value and original_unit as in the example above?

[aanderss]

This should be handled in a consistent way. if we take units from c99 attachment then the value needs to be read from there, too. Otherwise the conversion_method may be not consistent with the vales in the tables.

In this case the original_value is read from core and the original_units should be taken from core as well.

[ludwiglierhammer]

@aanderss: Just for clearification:

In this case, original_units is taken from the c99_data attachment for observation tables AT and SLP.
Because there are no values for air_temperature and sea_level_pressurein this attachment, original_valueis not set.
For the other observation tables, original_value is taken from coreand original_units is defined in the mapping json file.

For all observation tables, observation_value is taken from core and ``original_units` is defined in the mapping json files.

Question: Should original_valuebe taken from core for observation tables AT and SLP and original_units be defined in the mapping json files, as is the case for the other observation tables?

[ludwiglierhammer]

For now, we leave it as it is.
Further discussions in #72.