Not be8ing able to write_grid_mdv properly
MohitDahliya opened this issue · comments
- Py-ART version: 1.16.1
- Python version: 3.11.5
- Operating System: Windows
Description
I wanted to convert nc radr files into grid_mmdv files
While the conversion did happen but the filed of reflectivity was not there in the newly created grid_mdv file, only a few fields were there. I was expecting that all the files in the grid file would come into the mdv file
What I Did
import numpy as np
import pyart
radar = pyart.io.read('DLI190207055229-IMD-B.nc')
radar.fields.keys()
out >> dict_keys(['T', 'Z', 'V', 'W', 'ZDR', 'KDP', 'PHIDP', 'SQI', 'RHOHV', 'HCLASS'])
grid = pyart.map.grid_from_radars(radar, grid_shape=(41, 401, 401),
grid_limits=((0000.0, 20000.0), (-200000.0, 200000.0), (-200000.0, 200000.0)),
gridding_algo='map_gates_to_grid', copy_field_dtypes=True)
grid.fields.keys()
out >> dict_keys(['HCLASS', 'ZDR', 'W', 'SQI', 'PHIDP', 'RHOHV', 'T', 'Z', 'V', 'KDP', 'ROI'])
pyart.io.write_grid_mdv('output_grid.mdv', grid, mdv_field_names= None, field_write_order= None)
file = pyart.io.read_grid_mdv("output_grid.mdv")
file.fields.keys()
out >> dict_keys(['differential_reflectivity', 'cross_correlation_ratio', 'differential_phase', 'specific_differential_phase'])
@MohitDahliya - thanks for raising this issue, by default the mdv writer will only write the fields with mappings to mdv data
mdv_field_names (dict or None, optional) – Mapping between grid fields and MDV data type names. Field names mapped to None or with no mapping will be excluded from writing. If None, the same field names will be used.
https://arm-doe.github.io/pyart/API/generated/pyart.io.write_grid_mdv.html#pyart.io.write_grid_mdv
You would need to provide these mappings to write out the additional fields
How could I provide these mappings ?
I'm waiting for your response.
@MohitDahliya Sorry for the delay - I did not work on this during the weekend, and was attending the solar eclipse yesterday. We will take a look at this this week.
Okay.. Appreciate your response..
@MohitDahliya The field mapping is a dictionary, so you need to reference try to reference that SQI is normalized coherent power etc. Or when you read the grid try to set file_field_names = True so your using the fields in your dataset. The problem is the mapping here:
Line 1094 in 243e0c5
Does not match the fields you have in the raw dataset for mdv.
So you either need to provide a dictionary that says SQI: normalized_coherent_power as @mgrover1 stated, or try setting file_field_names = True to try to bypass trying to map nonmatching field names
@MohitDahliya - did that solution @zssherman work for you?
@MohitDahliya - we are following up here, did that solution work?
yes, it worked. I checked default_config.py and found that I needed to change the name of my reflectivity field from 'Z' to 'DBZ'. However, I think you people should change this fixed file name and instead focus on the long name of the field as not everyone would have the name of the reflectivity field as 'DBZ' in their radar file.
@mgrover1 I saw that you have also handled some issues related to CSU_Radartools. I am also facing some issues in running that code. Can you help me with that ?
yes, it worked. I checked default_config.py and found that I needed to change the name of my reflectivity field from 'Z' to 'DBZ'. However, I think you people should change this fixed file name and instead focus on the long name of the field as not everyone would have the name of the reflectivity field as 'DBZ' in their radar file.
Pull requests are always welcome!
@mgrover1 I saw that you have also handled some issues related to CSU_Radartools. I am also facing some issues in running that code. Can you help me with that ?
I would suggest opening an issue on the repository describing your issue. I am on travel this week, and likely would not be able to take a look until next week. One of the other developers there should be able to help.