add errors and explicit messages

pyam/core.py

@@ -48,7 +48,8 @@
     META_IDX,
     YEAR_IDX,
     IAMC_IDX,
-    SORT_IDX
+    SORT_IDX,
+    ILLEGAL_COLS
 )
 from pyam.read_ixmp import read_ix
 from pyam.timeseries import fill_series
@@ -618,8 +619,10 @@ def set_meta(self, meta, name=None, index=None):
         if (name or (hasattr(meta, 'name') and meta.name)) in [None, False]:
             raise ValueError('Must pass a name or use a named pd.Series')
         name = name or meta.name
-        if name in self.data.columns:
-            raise ValueError(f'A column `{name}` already exists in `data`!')
+        if name in self._data.index.names:
+            raise ValueError(f'Column {name} already exists in `data`!')
+        if name in ILLEGAL_COLS:
+            raise ValueError(f'Name {name} is illegal for meta indicators!')
 
         # check if meta has a valid index and use it for further workflow
         if hasattr(meta, 'index') and hasattr(meta.index, 'names') \

pyam/utils.py

@@ -25,6 +25,9 @@
 SORT_IDX = ['model', 'scenario', 'variable', 'year', 'region']
 LONG_IDX = IAMC_IDX + ['year']
 
+# illegal terms for data/meta column names to prevent attribute conflicts
+ILLEGAL_COLS = ['data', 'meta']
+
 # dictionary to translate column count to Excel column names
 NUMERIC_TO_STR = dict(zip(range(0, 702),
                           [i for i in string.ascii_uppercase]
@@ -133,7 +136,7 @@ def format_data(df, **kwargs):
         df.name = df.name or 'value'
         df = df.to_frame()
 
-    # Check for R-style year columns, converting where necessary
+    # check for R-style year columns, converting where necessary
     def convert_r_columns(c):
         try:
             first = c[0]
@@ -203,6 +206,14 @@ def convert_r_columns(c):
     if not list(df.index.names) == [None]:
         df.reset_index(inplace=True)
 
+    # check that there is no column in the timeseries data with reserved names
+    conflict_cols = [i for i in df.columns if i in ILLEGAL_COLS]
+    if conflict_cols:
+        msg = f'Column name {conflict_cols} is illegal for timeseries data.\n'
+        _args = ', '.join([f"{i}_alt='{i}'" for i in conflict_cols])
+        msg += f'Use `IamDataFrame(..., {_args})` to rename at initalization.'
+        raise ValueError(msg)
+
     # format columns to lower-case and check that all required columns exist
     if not set(IAMC_IDX).issubset(set(df.columns)):
         missing = list(set(IAMC_IDX) - set(df.columns))