ValueError: operands could not be broadcast together with shapes (5,) (10,)

[WieFel]

Describe the bug Hi, I am trying to use the pm_fao56 function to calculate the PET, but I am always getting the following error:

...
Cell In[41], [line 13](vscode-notebook-cell:?execution_count=41&line=13)
     [10] lat = 48.32829 * np.pi / 180  # Latitude of the meteorological station, converting from degrees to radians
     [11] elevation = 180  # meters above sea-level
---> [13] pet_df = pyet.pm_fao56(tmean, wind, rs=rs, elevation=elevation, lat=lat, rh=rh)
     [14] return pet_df

File .venv/lib/python3.11/site-packages/pyet/combination.py:626, in pm_fao56(tmean, wind, rs, rn, g, tmax, tmin, rhmax, rhmin, rh, pressure, elevation, lat, n, nn, rso, a, b, ea, albedo, kab, as1, bs1, clip_zero)
    [622] pressure, gamma, dlt, lambd, ea, es = _lambda_gamma_dlt_ea_es(
    [623]     elevation, pressure, tmean, tmax, tmin, rhmax, rhmin, rh, ea
    [624] )
    [625] gamma1 = gamma * (1 + 0.34 * wind)
--> [626] rn = calc_rad_net(
    [627]     tmean,
    [628]     rn,
    [629]     rs,
    [630]     lat,
    [631]     n,
    [632]     nn,
    [633]     tmax,
...
File .venv/lib/python3.11/site-packages/pandas/core/roperator.py:15, in rsub(left, right)
     [14] def rsub(left, right):
---> [15]     return right - left

ValueError: operands could not be broadcast together with shapes (5,) (10,)

Additionally, I am always seeing the following warning. I am not passing any timestamp into the pm_fao56 function. So I wonder where this warning can come from.

pyrealm/rad_utils.py:210: RuntimeWarning: '<' not supported between instances of 'Timestamp' and 'int', sort order is undefined for incomparable objects.

To Reproduce Steps to reproduce the behavior:

1. Use the following code:

   
   import pyet

meteo = pd.DataFrame({'tmean': [27.1 , 27.16, 27.22, 27.28, 27.34], 'rh': [42. , 41.6, 41.2, 40.8, 40.4], 'wind': [1.876, 1.876, 1.876, 1.876, 1.876], 'rs': [ 8.52768 , 10.062144, 11.596608, 13.131072, 14.665536]}) print(meteo) tmean, rh, wind, rs = [meteo[col] for col in meteo.columns]

lat = 48.32829 * np.pi / 180 # Latitude of the meteorological station, converting from degrees to radians elevation = 180 # meters above sea-level

pet_df = pyet.pm_fao56(tmean, wind, rs=rs, elevation=elevation, lat=lat, rh=rh)

where the meteo dataframe looks like this:

tmean rh wind rs 0 27.10 42.0 1.876 8.527680 1 27.16 41.6 1.876 10.062144 2 27.22 41.2 1.876 11.596608 3 27.28 40.8 1.876 13.131072 4 27.34 40.4 1.876 14.665536

2. See the error output

**Expected behavior**
The calculation should be executed without an error.

**Additional context**
Using Python `3.11.6` and pyet version `1.3.1`

[WieFel]

I tried to find the issue in the source code, and came to the point that in the function calc_rad_long https://github.com/pyet-org/pyet/blob/ff411787fbaeaa903b9966e28df41db9afb05a28/pyet/rad_utils.py#L202 the dataframe in rso is containing a datetime index, and that is apparently causing the shape of rs / rso to be 2x the expected shape. Both, the shapes of rs and rso are (5,).

rs looks like this:

0    0.0
1    0.0
2    0.0
3    0.0
4    0.0

rso looks like this:

1970-01-01 00:00:00.000000000    6.58243
1970-01-01 00:00:00.000000001    6.58243
1970-01-01 00:00:00.000000002    6.58243
1970-01-01 00:00:00.000000003    6.58243
1970-01-01 00:00:00.000000004    6.58243

However, the resulting shape of rs / rso is (10,).

Changing rs / rso to rs.values / rso.values solves the problem. But I am asking myself if that is the desired solution!?