id,node_id,number,title,user,state,locked,assignee,milestone,comments,created_at,updated_at,closed_at,author_association,active_lock_reason,draft,pull_request,body,reactions,performed_via_github_app,state_reason,repo,type
949649935,MDU6SXNzdWU5NDk2NDk5MzU=,5625,Add 'construct' method to Coarsen objects,18679148,closed,0,,,1,2021-07-21T12:20:39Z,2021-07-21T14:25:11Z,2021-07-21T14:25:11Z,NONE,,,,"**Is your feature request related to a problem? Please describe.**
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

Similarly to the 'construct' method for Rolling objects, I think it will make sense to have the same for Coarsen objects.

My use will be for my da:
da(lon, lat, time) of dimension (2000, 1600, 240)
I wish to coarsen the data by a factor 4 in lon and lat (Lon and Lat), but wish to keep the data in a new dimension, it would be:
da(Lon, Lat, time, samples) of dim (500, 400, 240, 16)

my aim is to do da.median(dim=['time','samples'])

**Describe the solution you'd like**
A clear and concise description of what you want to happen.

da.coarsen(lon=4, lat=4, , boundary=""trim"").construct('samples')

**Describe alternatives you've considered**
A clear and concise description of any alternative solutions or features you've considered.

I am using Rolling objects, but it increases the size of the matrix.

Many thanks for your work
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5625/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,completed,13221727,issue
700915153,MDU6SXNzdWU3MDA5MTUxNTM=,4421,circular + linear stats on a same dataset,18679148,open,0,,,1,2020-09-14T09:03:24Z,2020-11-01T13:44:26Z,,NONE,,,,"<!-- Please do a quick search of existing issues to make sure that this has not been asked before. -->

**Is your feature request related to a problem? Please describe.**
I would like to calculate the mean on all variables of my dataset which is easy and straightforward with xarray, for example:
`ds.mean(dim='time)`
The issue is I have some circular quantities in this dataset (for example wind direction). I can use script.stats.circmean/std, but I have to discriminate myself when I have to use one or the other function.

**Describe the solution you'd like**
I would like to add a parameter in the description of my variable mentioning it is circular. Every time I apply a function on the full dataset, it uses the circular equivalent if it exists.

There is perhaps some tricks to do it, but I didn't find any so far. Thank you.
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/4421/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,issue
663649344,MDU6SXNzdWU2NjM2NDkzNDQ=,4246,combine_by_coords; proposition for a new option for combine_attrs = 'dim' ,18679148,closed,0,,,2,2020-07-22T10:22:32Z,2020-07-23T18:46:51Z,2020-07-23T18:46:51Z,NONE,,,,"I am combining a list of snapshots having all the same geometry but with different time. Some time information appears in the attributes. I can 'drop' it, but I would prefer keep it and add it using a define dimension (in my case, in time).
I believe for v0.15.1 (the default was to drop it with the default compat='no_conflicts'), I think of this because I got an error on the default combine_attrs = 'no_conflicts' on v0.16.0.

I would like an option of type combine_attrs = {dim: 'time'} or even better only combine_attrs = 'dim' and somehow it finds which dimension it should use.

Thanks","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/4246/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,completed,13221727,issue
663769801,MDU6SXNzdWU2NjM3Njk4MDE=,4247,plot/utils get_axis cannot use subplot_kws with existing ax,18679148,closed,0,,,2,2020-07-22T13:38:04Z,2020-07-22T14:59:16Z,2020-07-22T14:59:16Z,NONE,,,,"I get the following error when I move from xarray 0.15.1 to 0.16.0 ""cannot use subplot_kws with existing ax"". I don't really understand why this error is raised, but as far as I use it (see below), I feel I need to use both ax and subplot_kws.
Please find below a minimum case:
```python
import numpy as np
import pandas as pd
import xarray as xr
import matplotlib.pyplot as plt
import cartopy.crs as ccrs

x = np.arange(0,3)
y = np.arange(0,5)
z = np.arange(0,2)
xx, yy, zz = np.meshgrid(x, y, z)

lat = 40 + xx*0.1 + yy*1
lon = 5 + xx*1 + yy*0.1 + zz*5
data3 = xx + yy + 5*zz
data4 = np.stack([data3, data3*2, data3*4, data3*8])

ds = xr.Dataset({'data': (['time','x','y','z'], data4)},
               coords={'time': pd.date_range(""2020-01-01"", periods=4),
                       'lon': (['x','y','z'], lon),
                       'lat': (['x','y','z'], lat)}
               )
```

So I have lon/lat maps with two swaths (z-dimension).
I want to plot both swaths on the same map, and as function of time 
```python
plt.figure()
for ss in z:
    ds.data.isel(time=0, z=ss).plot(x='lon', y='lat', add_colorbar=False, vmin=0, vmax=12)
    
plt.figure()
g = ds.data.isel(z=0).plot(x='lon', y='lat', col='time')
for cc, ax in enumerate(g.axes.flat):
    for ss in z:
        ds.data.isel(time=cc, z=ss).plot(x='lon', y='lat', add_colorbar=False, vmin=0, vmax=50, ax=ax)
```

All is relatively fine, the issue is when I want to project it on a map using Cartopy:
Still fine with only one map:
```python
plt.figure()
for ss in z:
    ds.data.isel(time=0, z=ss).plot(x='lon', y='lat', add_colorbar=False, vmin=0, vmax=12,
    transform=ccrs.PlateCarree(), subplot_kws={'projection': ccrs.Mercator()})
```

but the ""ValueError: cannot use subplot_kws with existing ax"" is raised when I try with the facets plot:
```python
plt.figure()
g = ds.data.isel(z=0).plot(x='lon', y='lat', col='time', 
transform=ccrs.PlateCarree(), subplot_kws={'projection': ccrs.Mercator()})
for cc, ax in enumerate(g.axes.flat):
    for ss in z:
        ds.data.isel(time=cc, z=ss).plot(x='lon', y='lat', add_colorbar=False, vmin=0, vmax=50, 
       ax=ax, transform=ccrs.PlateCarree(), subplot_kws={'projection': ccrs.Mercator()})
```

On a side note, I don't understand why I am loosing one pixel per row and column using the cartopy projection.
<details><summary>Output of <tt>xr.show_versions()</tt></summary>

<!-- Paste the output here xr.show_versions() here -->
INSTALLED VERSIONS
------------------
commit: None
python: 3.7.6 (default, Jan  8 2020, 13:42:34) 
[Clang 4.0.1 (tags/RELEASE_401/final)]
python-bits: 64
OS: Darwin
OS-release: 19.5.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_GB.UTF-8
LOCALE: en_GB.UTF-8
libhdf5: 1.10.4
libnetcdf: 4.6.1

xarray: 0.16.0
pandas: 1.0.1
numpy: 1.18.1
scipy: 1.4.1
netCDF4: 1.4.2
pydap: None
h5netcdf: None
h5py: None
Nio: None
zarr: None
cftime: 1.0.4.2
nc_time_axis: None
PseudoNetCDF: None
rasterio: None
cfgrib: None
iris: None
bottleneck: None
dask: 2.10.1
distributed: 2.10.0
matplotlib: 3.1.3
cartopy: 0.17.0
seaborn: None
numbagg: None
pint: None
setuptools: 45.2.0.post20200210
pip: 20.0.2
conda: None
pytest: None
IPython: 7.12.0
sphinx: None

</details>
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/4247/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,completed,13221727,issue