Module "skill_metrics" in WP5

[paolocristofanelli]

Hi @malmans2

I'm trying to run [trend_pr_cmip6.ipynb] but it looks that "skill_metrics" is not mounted on WP5.

The workflow looks very similar to the analysis done in trend_pr_cmip6.ipynb. In that notebook the trends are computed using a package named xarrayMannKendall. However, it's not available on the wp5 environment. Let me know if you need it on the VM.

Originally posted by @malmans2 in https://github.com/bopen/c3s-eqc-toolbox-template/issues/52#issuecomment-1540405951

[malmans2]

Yes, that's a notebook for wp4. Are you going to need it skill_metrics for your notebook (I think it's used to plot taylor diagrams)? If yes, I'll install it on wp5 as well.

[malmans2]

(If you don't need taylor diagrams, you should be able to comment out import skill_metrics , and run the whole notebook with exception of the last cell)

[malmans2]

Also, if you want to take advantage of the cache, you should probably run it as a wp4 user.

[paolocristofanelli]

I commented out import skill_metrics but I got the follwing error in "Compute spatial weighted trends".

AttributeError Traceback (most recent call last) Cell In[9], line 14 11 request_model = request_sim 12 request_model[1]["model"] = model 13 datasets.append( ---> 14 download.download_and_transform( 15 *request_model, 16 chunks=chunks, 17 transform_func=spatial_weighted_trends, 18 transform_func_kwargs={"model": model, **transform_func_kwargs}, 19 transform_chunks=False, 20 ) 21 ) 23 # Combine and add ensemble 24 ds_spatial_weighted = xr.concat(datasets, "model")

File /data/common/mambaforge/envs/wp5/lib/python3.10/site-packages/c3s_eqc_automatic_quality_control/download.py:528, in download_and_transform(collection_id, requests, chunks, split_all, transform_func, transform_func_kwargs, transform_chunks, n_jobs, open_mfdataset_kwargs) 525 ds = xr.open_mfdataset(sources, open_mfdataset_kwargs) 526 else: 527 # Cache final dataset transformed --> 528 ds = func(request_list=request_list) 530 ds.attrs.pop("coordinates", None) # Previously added to guarantee roundtrip 531 return ds

File /data/common/mambaforge/envs/wp5/lib/python3.10/site-packages/cacholote/cache.py:86, in cacheable..wrapper(*args, *kwargs) 83 warnings.warn(str(ex), UserWarning) 84 clean._delete_cache_entry(session, cache_entry) ---> 86 result = func(args, **kwargs) 87 cache_entry = database.CacheEntry( 88 key=hexdigest, 89 expiration=settings.expiration, 90 tag=settings.tag, 91 ) 92 try:

File /data/common/mambaforge/envs/wp5/lib/python3.10/site-packages/c3s_eqc_automatic_quality_control/download.py:433, in _download_and_transform_requests(collection_id, request_list, transform_func, transform_func_kwargs, open_mfdataset_kwargs) 430 ds = xr.open_mfdataset(sources, open_mfdataset_kwargs) 432 if transform_func is not None: --> 433 ds = transform_func(ds, **transform_func_kwargs) 434 if not isinstance(ds, xr.Dataset): 435 raise TypeError( 436 f"transform_func must return a xr.Dataset, while it returned {type(ds)}" 437 )

Cell In[8], line 37, in spatial_weighted_trends(ds, model, clima_year_start, clima_year_stop) 35 da = ds["precipitation"] 36 else: ---> 37 da = diagnostics.spatial_weighted_mean(get_annual_precipitation(ds, model)) 38 anoma = compute_anomaly(da, clima_year_start, clima_year_stop) 40 # Compute anomaly trends

Cell In[8], line 9, in get_annual_precipitation(ds, model) 6 varname = "mtpr" if model == "ERA5" else "pr" 8 # Annual weighted mean ----> 9 da = diagnostics.annual_weighted_mean_timeseries(ds[varname]) 11 # Convert units 12 with xr.set_options(keep_attrs=True):

AttributeError: module 'c3s_eqc_automatic_quality_control.diagnostics' has no attribute 'annual_weighted_mean_timeseries'

[malmans2]

I think you are using an old version of the notebook. Make sure you use the latest: https://github.com/bopen/c3s-eqc-toolbox-template/blob/main/notebooks/wp4/trend_pr_cmip6.ipynb

[malmans2]

I think we can close this? Feel free to re-open though.