How to loop over a list of statistic type and apply them on an xarray.DataArray?

Question

I need to compute a list statistics through time on a xarray.DataArray ans store them in a xarray.Dataset:

import xarray as xr
import numpy as np
import pandas as pd

np.random.seed(1234)
da = xr.DataArray(data=np.random.rand(4,5,10),
                  dims=["lon", "lat","time"],
                  coords={"lon": np.random.uniform(low=-90, high=90, size=4),
                          "lat": np.random.uniform(low=-90, high=90, size=5),
                          "time": pd.date_range(start="2021-01-01", freq="D", periods=10)})
da

first = True
if first:
    first = False
    ds = da.min(dim=['time']).to_dataset(name = "min")
else:
    ds = ds.merge(da.min(dim=['time']).to_dataset(name = "min"))
ds = ds.merge(da.max(dim=['time']).to_dataset(name = "max"))
ds = ds.merge(da.median(dim=['time']).to_dataset(name = "median"))
ds = ds.merge(da.mean(dim=['time']).to_dataset(name = "mean"))
ds = ds.merge(da.std(dim=['time']).to_dataset(name = "std"))

ds

As I need to frequently change the statistics to apply, I tried to use a list of statistics and loop through it:

stats = ('min', 'max', 'median', 'mean', 'std')

first = True
for stat in stats:
    if first:
        first = False
        ds = da.vars()['stat'](dim=['time']).to_dataset(name = vars()['stat'])
    else:
        ds = ds.merge(da.vars()['stat'](dim=['time']).to_dataset(name = vars()['stat']))

But I get an Error AttributeError: 'DataArray' object has no attribute 'vars'when trying to retrieve and apply the statistic type.

Thanks for any hint you could provide.

Answer 1

Thanks to @michael-delgado for informing me functions can as well be listed. Here is the answer I came through:

stats = [np.nanmin, np.nanmax, np.nanmedian, np.nanmean, np.nanstd]

first = True
for i in range(0, len(stats)):
    da_stat = xr.DataArray(stats[i](a = da, axis = 2), dims = ['lon', 'lat'])
    ds_stat = da_stat.assign_coords(lon = da.lon.values,
                                    lat = da.lat.values).to_dataset(name = stats[i].__name__)
    if first:
        first = False
        ds = ds_stat
    else:
        ds = ds.merge(ds_stat)

ds

The only downside with this solution, is I was not able to use xarray.DataArray statistic function, and had to replace them with numpy statistic functions, which double processing time.