Bugfix for non-empty invisible header in 'meta' sheet (#508)

RELEASE_NOTES.md

@@ -8,6 +8,7 @@ starting with `data` will be parsed for timeseries data.
 
 ## Individual updates
 
+- [#508](https://github.com/IAMconsortium/pyam/pull/508) Bugfix for non-empty but invisible header and no rows in 'meta' sheet
 - [#502](https://github.com/IAMconsortium/pyam/pull/502) Switch to Black code style
 - [#499](https://github.com/IAMconsortium/pyam/pull/499) Implement `order` feature in line plot 
 - [#497](https://github.com/IAMconsortium/pyam/pull/497) Add a module for reading data from the UNFCCC Data Inventory 

pyam/utils.py

@@ -156,11 +156,13 @@ def read_pandas(path, sheet_name="data*", *args, **kwargs):
             df = pd.read_excel(path, *args, **kwargs)
 
         # remove unnamed and empty columns, and rows were all values are nan
-        empty_cols = [
-            c
-            for c in df.columns
-            if str(c).startswith("Unnamed: ") and all(np.isnan(df[c]))
-        ]
+        def is_empty(name, s):
+            if str(name).startswith("Unnamed: "):
+                if len(s) == 0 or all(np.isnan(s)):
+                    return True
+            return False
+
+        empty_cols = [c for c in df.columns if is_empty(c, df[c])]
         return df.drop(columns=empty_cols).dropna(axis=0, how="all")
 
 

tests/test_io.py

@@ -171,7 +171,7 @@ def test_load_meta_wrong_index(test_df_year, tmpdir):
 
 
 def test_load_meta_empty_rows(test_df_year, tmpdir):
-    """Loading empty meta table (columns but no rows) from an Excel file"""
+    """Loading empty meta table (columns but no rows) from xlsx file"""
     exp = test_df_year.copy()  # loading empty file has no effect
 
     # write empty meta frame to file, then load to the IamDataFrame
@@ -182,6 +182,13 @@ def test_load_meta_empty_rows(test_df_year, tmpdir):
     assert_iamframe_equal(test_df_year, exp)
 
 
+def test_load_meta_empty(test_pd_df):
+    """Initializing from xlsx where 'meta' has no rows and non-empty invisible header"""
+    obs = IamDataFrame(TEST_DATA_DIR / "empty_meta_sheet.xlsx")
+    exp = IamDataFrame(test_pd_df)
+    assert_iamframe_equal(obs, exp)
+
+
 def test_load_ssp_database_downloaded_file(test_pd_df):
     exp = IamDataFrame(test_pd_df).filter(**FILTER_ARGS).as_pandas()
     file = TEST_DATA_DIR / "test_SSP_database_raw_download.xlsx"