Compute the mean, median, modes, counts from an xarray DataArrayResample object, returns a pandas dataframe

resampled_stats.py

# requires: numpy, pandas, scipy.stats, xarray

def compute_modes(resampled, n=0):
    
    '''Given a resampled xarray object (xarray.core.resample.DataArrayResample),
       compute the modes ( rounding values with np.round(_,n) ), 
       return a pandas dataframe with the modes, counts, and groups (datetime)'''
    
    # Compute modes
    resampled_ModeResults = [stats.mode(np.round(x[1],n)) for x in resampled]
    
    # Get all the modes
    resampled_modes = [x.mode for x in resampled_ModeResults]
    # Reshape result into 1D array
    resampled_modes = np.array(resampled_modes).reshape(len(resampled))
    
    # Get the count for each mode
    resampled_counts = [x.count for x in resampled_ModeResults]
    # Reshape result into 1D array
    resampled_counts = np.array(resampled_counts).reshape(len(resampled))
    
    # Get the group (datetime) for each mode
    resampled_groups = np.array( list(resampled.groups.keys()) )
    
    # Create pandas dataframe
    d = {'modes': resampled_modes, 'counts': resampled_counts}
    df = pd.DataFrame(data=d, index=resampled_groups)
    
    return df

def resampled_stats(resampled, n=0, q=(.25,.75)):
    
    '''Given a resampled xarray object (xarray.core.resample.DataArrayResample),
    return a pandas dataframe with the mean, median, modes, counts, and groups (datetime)'''
    
    # Compute modes
    resampled_modes = compute_modes(resampled, n)
    
    # Compute and add mean, medians, modes, counts to a dataframe
    d = {'means': resampled.mean().values, 
         'medians': resampled.median().values, 
         'modes': resampled_modes.modes.values,
         'counts': resampled_modes.counts.values,
         'maxs': resampled.max().values,
         'mins': resampled.min().values,
         'ranges': resampled.max().values - resampled.min().values,
         'stds': resampled.std().values,
         'qUpper': resampled.quantile(q[1]).values,
         'qLower': resampled.quantile(q[0]).values,}
    
    df = pd.DataFrame(data=d, index=resampled_modes.index)
    
    return df