data-request-tools

Tools to manage a (CORDEX) data request in csv format. The main task of this package is to create CORDEX-CMIP6 create tables) from a data request table in csv format. In general, this package should also be useful to create other tables from other data requests.

API Example

This tools helps to convert a CMIP data request in csv format into a number of CMOR tables in json format.

import data_request as dr
import pandas as pd

table = "https://raw.githubusercontent.com/WCRP-CORDEX/data-request-table/refs/heads/main/cmor-table/datasets.csv"

df = pd.read_csv(table).fillna("")
cmor_tables = dr.create_cmor_tables(df)
cmor_tables.keys()

gives

dict_keys(['1hr', '6hr', 'day', 'fx', 'mon'])

The tables can be written to CMOR json tables using:

for t in cmor_tables.values():
    dr.table_to_json(t, table_prefix="CORDEX-CMIP6")

which creates a tables based on the table_id in the header:

writing: ./CORDEX-CMIP6_1hr.json
writing: ./CORDEX-CMIP6_6hr.json
writing: ./CORDEX-CMIP6_day.json
writing: ./CORDEX-CMIP6_fx.json
writing: ./CORDEX-CMIP6_mon.json

The coordinates table can also be created by searching through the dimensions column:

coords = dr.create_coordinate_table(df)

results in

found: ['height100m', 'height10m', 'height150m', 'height200m', 'height250m', 'height2m', 'height300m', 'height50m', 'latitude', 'longitude', 'p10', 'p100', 'p1000', 'p150', 'p20', 'p200', 'p250', 'p30', 'p300', 'p400', 'p50', 'p500', 'p600', 'p70', 'p700', 'p750', 'p850', 'p925', 'sdepth', 'time', 'time1']

and can be written using:

dr.table_to_json(coords, "CORDEX-CMIP6", table_id="coordinate")

Command line

You can also use the command create-cmor-tables in a terminal to create cmor tables from anything that pandas.read_csv can understand, e.g.

create-cmor-tables https://raw.githubusercontent.com/WCRP-CORDEX/data-request-table/refs/heads/main/cmor-table/datasets.csv --coords

will do the same as above.