ENH: Return column information in add_constant

Return information about constant columns in add_constant

closes #6756

statsmodels/tools/tools.py

@@ -317,7 +317,8 @@ def add_constant(data, prepend=True, has_constant='skip'):
 
     # Special case for NumPy
     x = np.asanyarray(data)
-    if x.ndim == 1:
+    ndim = x.ndim
+    if ndim == 1:
         x = x[:, None]
     elif x.ndim > 2:
         raise ValueError('Only implemented for 2-dimensional arrays')
@@ -328,7 +329,12 @@ def add_constant(data, prepend=True, has_constant='skip'):
         if has_constant == 'skip':
             return x
         elif has_constant == 'raise':
-            raise ValueError("data already contains a constant")
+            if ndim == 1:
+                raise ValueError("data is constant.")
+            else:
+                columns = np.arange(x.shape[1])
+                cols = ",".join([str(c) for c in columns[is_nonzero_const]])
+                raise ValueError(f"Column(s) {cols} are constant.")
 
     x = [np.ones(x.shape[0]), x]
     x = x if prepend else x[::-1]

statsmodels/tsa/tests/test_arima.py

@@ -2580,7 +2580,7 @@ def test_constant_column_trend():
     model = ARIMA(endog=endog, order=(1, 1, 0), exog=exog)
 
     # Fitting with a constant and constant exog raises because of colinearity
-    with pytest.raises(ValueError, match="x contains a constant"):
+    with pytest.raises(ValueError, match="x contains one or more constant"):
         model.fit(trend="c", disp=-1)
 
     # FIXME: calling model.fit(trend="nc") raises for orthogonal reasons

statsmodels/tsa/tsatools.py

@@ -116,8 +116,20 @@ def safe_is_const(s):
 
         if np.any(col_const):
             if has_constant == 'raise':
-                msg = "x contains a constant. Adding a constant with " \
-                      "trend='{0}' is not allowed.".format(trend)
+                if x.ndim == 1:
+                    base_err = "x is constant."
+                else:
+                    columns = np.arange(x.shape[1])[col_const]
+                    if isinstance(x, pd.DataFrame):
+                        columns = x.columns
+                    const_cols = ", ".join([str(c) for c in columns])
+                    base_err = (
+                        "x contains one or more constant columns. Column(s) "
+                        f"{const_cols} are constant."
+                    )
+                msg = (
+                    f"{base_err} Adding a constant with trend='{trend}' is not allowed."
+                )
                 raise ValueError(msg)
             elif has_constant == 'skip':
                 columns = columns[1:]