Make SGI applicable to DataFrame and add Dropna (#760)

pastas/stats/sgi.py

@@ -1,7 +1,7 @@
 """This module contains methods to compute the Standardized Groundwater Index."""
 
 from numpy import linspace
-from pandas import Series
+from pandas import DataFrame, Series
 from scipy.stats import norm
 
 
@@ -11,22 +11,35 @@ def sgi(series: Series) -> Series:
 
     Parameters
     ----------
-    series: pandas.Series
+    series: pandas.Series or Pandas.DataFrame
 
     Returns
     -------
-    sgi_series: pandas.Series
+    sgi_series: pandas.Series or Pandas.DataFrame
         Pandas time series of the groundwater levels. Time series index should be a
         pandas DatetimeIndex.
+
+    Notes
+    -----
+    The Standardized Groundwater Index (SGI) is a non-parametric method to
+    standardize groundwater levels. The SGI is calculated for each month
+    separately. The SGI is a dimensionless index and is used to compare
+    groundwater levels across different wells. It is generally recommended to resample
+    the data to a monthly time series before computing the SGI.
+
     """
-    series = series.copy()  # Create a copy to ensure series is untouched.
-
-    # Loop over the months
-    for month in range(1, 13):
-        data = series[series.index.month == month]
-        n = data.size  # Number of observations
-        pmin = 1 / (2 * n)
-        pmax = 1 - pmin
-        sgi_values = norm.ppf(linspace(pmin, pmax, n))
-        series.loc[data.sort_values().index] = sgi_values
+    if isinstance(series, DataFrame):
+        series = series.apply(sgi)
+    elif isinstance(series, Series):
+        series = series.dropna().copy()  # Create a copy to ensure series is untouched.
+
+        # Loop over the months
+        for month in range(1, 13):
+            data = series[series.index.month == month]
+            n = data.size  # Number of observations
+            pmin = 1 / (2 * n)
+            pmax = 1 - pmin
+            sgi_values = norm.ppf(linspace(pmin, pmax, n))
+            series.loc[data.sort_values().index] = sgi_values
+
     return series