Error: Changing months combination in xarray season for India Region #6012

HARSHBARDHANK94 · 2021-11-22T06:08:05Z

HARSHBARDHANK94
Nov 22, 2021

In xarray seasonal grouping includes combination of months : DJF, MAM, JJA, SON, but Indian Mainland seasonal climatology include: JJAS (monsoon), ON (post-monsoon), MAM (pre-monsoon) and DJF (winter). For the same I tried the following way to script:

def is_jjas(month):
    return (month >= 6) & (month <= 9)
JJAS = ds.sel (time=is_jjas (ds['time.month']))

def is_on(month):
    return (month >= 10) & (month <= 11)
ON = ds.sel (time=is_on (ds['time.month']))

def is_mam(month):
    return (month >= 3) & (month <= 5)
MAM = ds.sel (time=is_mam (ds['time.month']))

def is_djf(month):
    return (month == 12) & (month == 1) & (month == 2)
DJF = ds.sel (time=is_djf (ds['time.month'][2:-1]))   # [2:-1] slicing is used to remove Jan, Feb from the first and Dec from the last year in the datasets

The dataset (ds) has time dim of shape (216, ) i.e., monthly for 18 years, after slicing [2:-1] to remove the Jan, Feb from the first and Dec from the last year in the datasets, the shape reduce to (213, ).
That is why I got an error:

Traceback (most recent call last):

  File "C:\Users\HARSHB~1\AppData\Local\Temp/ipykernel_1252/1680323816.py", line 4, in <module>
    DJF = ds.sel (time=is_djf (ds['time.month'][2:-1]))

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataarray.py", line 1315, in sel
    ds = self._to_temp_dataset().sel(

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2477, in sel
    result = self.isel(indexers=pos_indexers, drop=drop)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2325, in isel
    return self._isel_fancy(indexers, drop=drop, missing_dims=missing_dims)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2383, in _isel_fancy
    new_var, new_index = isel_variable_and_index(

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\indexes.py", line 507, in isel_variable_and_index
    new_variable = variable.isel(indexers)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 1156, in isel
    return self[key]

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 776, in __getitem__
    dims, indexer, new_order = self._broadcast_indexes(key)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 619, in _broadcast_indexes
    self._validate_indexers(key)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 660, in _validate_indexers
    raise IndexError(

IndexError: Boolean array size 213 is used to index array with shape (216,).

Looking hopefully for any constructive remark or help from the community. Thanks.........

Answered by keewis

Nov 22, 2021

if you look at the implementation of time.season, it will translate months to season labels here:

xarray/xarray/core/accessor_dt.py

Lines 15 to 20 in 7a201de

     def _season_from_months(months):  
   """Compute season (DJF, MAM, JJA, SON) from month ordinal"""  
   # TODO: Move "season" accessor upstream into pandas  
   seasons = np.array(["DJF", "MAM", "JJA", "SON"])  
   months = np.asarray(months)  
   return seasons[(months // 3) % 4]  

 

Since your groups are not that different, you can take the expression (months // 3) % 4 and change the group for September (9):

def indian_mainland_seasons(months):
    seasons = np.array(["DJF", "MAM", "JJAS", "ON"])
    months = np.asarray(

View full answer

max-sixty · 2021-11-22T09:02:22Z

max-sixty
Nov 22, 2021
Maintainer

Quick response: does moving the index outside the selection help?

ds.sel (time=is_djf (ds)) ['time.month'][2:-1]

1 reply

HARSHBARDHANK94 Nov 22, 2021
Author

Traceback (most recent call last):

  File "C:\Users\HARSHB~1\AppData\Local\Temp/ipykernel_1252/2036856586.py", line 4, in <module>
   ds.sel (time=is_djf (ds)) ['time.month'][2:-1]

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataarray.py", line 1315, in sel
    ds = self._to_temp_dataset().sel(

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2477, in sel
    result = self.isel(indexers=pos_indexers, drop=drop)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2325, in isel
    return self._isel_fancy(indexers, drop=drop, missing_dims=missing_dims)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\dataset.py", line 2383, in _isel_fancy
    new_var, new_index = isel_variable_and_index(

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\indexes.py", line 507, in isel_variable_and_index
    new_variable = variable.isel(indexers)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 1156, in isel
    return self[key]

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 776, in __getitem__
    dims, indexer, new_order = self._broadcast_indexes(key)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 619, in _broadcast_indexes
    self._validate_indexers(key)

  File "C:\Users\HARSHBARDHAN\anaconda3\lib\site-packages\xarray\core\variable.py", line 665, in _validate_indexers
    raise IndexError(

IndexError: 3-dimensional boolean indexing is not supported.

HARSHBARDHANK94 · 2021-11-22T10:14:15Z

HARSHBARDHANK94
Nov 22, 2021
Author

Quick response: does moving the index outside the selection help?

ds.sel (time=is_djf (ds)) ['time.month'][2:-1]

not worked!!

0 replies

keewis · 2021-11-22T12:50:32Z

keewis
Nov 22, 2021
Maintainer

if you look at the implementation of time.season, it will translate months to season labels here:

xarray/xarray/core/accessor_dt.py

Lines 15 to 20 in 7a201de

    
           def _season_from_months(months): 
        
               """Compute season (DJF, MAM, JJA, SON) from month ordinal""" 
        
               # TODO: Move "season" accessor upstream into pandas 
        
               seasons = np.array(["DJF", "MAM", "JJA", "SON"]) 
        
               months = np.asarray(months) 
        
               return seasons[(months // 3) % 4]

Since your groups are not that different, you can take the expression (months // 3) % 4 and change the group for September (9):

def indian_mainland_seasons(months):
    seasons = np.array(["DJF", "MAM", "JJAS", "ON"])
    months = np.asarray(months)
    return seasons[np.where(months == 9, 2, (months // 3) % 4)]

This assumes numpy-like arrays, so you might need to use apply_ufunc to apply it to xarray objects. For example:

In [6]: ds = xr.tutorial.open_dataset("rasm")
   ...: 
   ...: def indian_mainland_seasons(arr):
   ...:     def compute(months):
   ...:         seasons = np.array(["DJF", "MAM", "JJAS", "ON"])
   ...:         months = np.asarray(months)
   ...:         return seasons[np.where(months == 9, 2, (months // 3) % 4)]
   ...: 
   ...:     return xr.apply_ufunc(
   ...:         compute,
   ...:         arr.dt.month,
   ...:     )
   ...: 
   ...: indian_mainland_seasons(ds.time)
Out[6]: 
<xarray.DataArray 'month' (time: 36)>
array(['JJAS', 'ON', 'ON', 'DJF', 'DJF', 'DJF', 'MAM', 'MAM', 'MAM',
       'JJAS', 'JJAS', 'JJAS', 'JJAS', 'ON', 'ON', 'DJF', 'DJF', 'DJF',
       'MAM', 'MAM', 'MAM', 'JJAS', 'JJAS', 'JJAS', 'JJAS', 'ON', 'ON',
       'DJF', 'DJF', 'DJF', 'MAM', 'MAM', 'MAM', 'JJAS', 'JJAS', 'JJAS'],
      dtype='<U4')
Coordinates:
  * time     (time) object 1980-09-16 12:00:00 ... 1983-08-17 00:00:00

1 reply

HARSHBARDHANK94 Nov 23, 2021
Author

Nice!! Perfectly worked 🙏 Thanks for quick support. It will helps many one , working in South-East Asian region seasonal climatology.

fmaussion · 2021-11-22T18:35:14Z

fmaussion
Nov 22, 2021
Maintainer

Also return (month == 12) & (month == 1) & (month == 2) should be return (month == 12) | (month == 1) | (month == 2)

0 replies

HARSHBARDHANK94 · 2021-11-23T06:31:18Z

HARSHBARDHANK94
Nov 23, 2021
Author

Also return (month == 12) & (month == 1) & (month == 2) should be return (month == 12) | (month == 1) | (month == 2)

Thanks, it also worked perfectly, in case one by one month is selected. 👍🙌

0 replies

dcherian · 2025-05-07T17:17:55Z

dcherian
May 7, 2025
Maintainer

On latest Xarray (>2025.04.0) you can do

import xarray as xr
from xarray.groupers import SeasonGrouper

ds = xr.tutorial.open_dataset("air_temperature")

ds.groupby(time=SeasonGrouper(["JJAS", "ON", "MAM", "DJF"])).mean()

https://docs.xarray.dev/en/latest/user-guide/time-series.html#handling-seasons

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Error: Changing months combination in xarray season for India Region #6012

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 6 comments 2 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Select a reply

	def _season_from_months(months):
	"""Compute season (DJF, MAM, JJA, SON) from month ordinal"""
	# TODO: Move "season" accessor upstream into pandas
	seasons = np.array(["DJF", "MAM", "JJA", "SON"])
	months = np.asarray(months)
	return seasons[(months // 3) % 4]

Error: Changing months combination in xarray season for India Region #6012

HARSHBARDHANK94 Nov 22, 2021

Replies: 6 comments · 2 replies

max-sixty Nov 22, 2021 Maintainer

HARSHBARDHANK94 Nov 22, 2021 Author

HARSHBARDHANK94 Nov 22, 2021 Author

keewis Nov 22, 2021 Maintainer

HARSHBARDHANK94 Nov 23, 2021 Author

fmaussion Nov 22, 2021 Maintainer

HARSHBARDHANK94 Nov 23, 2021 Author

dcherian May 7, 2025 Maintainer

HARSHBARDHANK94
Nov 22, 2021

Replies: 6 comments 2 replies

max-sixty
Nov 22, 2021
Maintainer

HARSHBARDHANK94 Nov 22, 2021
Author

HARSHBARDHANK94
Nov 22, 2021
Author

keewis
Nov 22, 2021
Maintainer

HARSHBARDHANK94 Nov 23, 2021
Author

fmaussion
Nov 22, 2021
Maintainer

HARSHBARDHANK94
Nov 23, 2021
Author

dcherian
May 7, 2025
Maintainer