Merge pull request #28 from NREL/robust_set_fix

Bug fix and change README to point to pypi for installation.

CHANGES.txt

@@ -1,3 +1,4 @@
+v1.0.4, 03/02/18 -- make sure set 'Value' column fix works even when dims are all named '*'
 v1.0.3, 02/15/18 -- fix up set 'Value' column when write to GDX
 v1.0.2, 01/19/18 -- instructions and additional filepaths for running pytest on 
                     installed package

README.md

@@ -1,5 +1,5 @@
-gdx-pandas
-==========
+# gdx-pandas
+[![PyPI](https://img.shields.io/pypi/v/gdxpds.svg)](https://pypi.python.org/pypi/gdxpds/)
 
 Python package to translate between gdx (GAMS data) and pandas. 
 
@@ -145,16 +145,24 @@ with gdxpds.gdx.GdxFile() as gdx:
 ### Get the Latest Package
 
 ```bash
-pip install git+https://github.com/NREL/gdx-pandas.git@master
+pip install gdxpds
 ```
 
-or 
+or
 
 ```bash
-pip install git+https://github.com/NREL/[email protected].3
+pip install git+https://github.com/NREL/[email protected].4
 ```
 
-Versions are listed at https://github.com/NREL/gdx-pandas/releases.
+or
+
+```bash
+pip install git+https://github.com/NREL/gdx-pandas.git@master
+```
+
+
+Versions are listed at [pypi](https://pypi.python.org/pypi/gdxpds/) and 
+https://github.com/NREL/gdx-pandas/releases.
 
 After installation, you can test the package using pytest:
 

gdxpds/__init__.py

@@ -43,7 +43,7 @@
 '''
 
 
-__version__ = '1.0.3'
+__version__ = '1.0.4'
 
 import logging
 import os

gdxpds/gdx.py

@@ -137,6 +137,16 @@ def gdx_val_equal(val1,val2,gdxf):
         return True
     return val1 == val2
 
+def replace_df_column(df,colname,new_col):
+    # This weirdness with replacing columns is a workaround for 
+    # pandas crashing on multiple columns called '*'.
+    cols = df.columns
+    tmpcols = [col if col != '*' else 'aaa' for col in cols ]
+    df.columns = tmpcols
+    df[colname] = new_col
+    df.columns = cols
+    return
+
 
 class GdxError(Error):
     def __init__(self, H, msg):
@@ -657,20 +667,15 @@ def dataframe(self, data):
             self._dataframe = pds.DataFrame(data,columns=self.dims + self.value_col_names)
 
         if self.data_type == GamsDataType.Set:
+            logger.debug(self._dataframe.head())
             self._fixup_set_value()
         return
 
     def _init_dataframe(self):
         self._dataframe = pds.DataFrame([],columns=self.dims + self.value_col_names)
         if self.data_type == GamsDataType.Set:
             colname = self._dataframe.columns[-1]
-            # This weirdness with replacing columns is a workaround for 
-            # pandas crashing on multiple columns called '*'.
-            cols = self._dataframe.columns
-            tmpcols = [col if col != '*' else 'aaa' for col in cols ]
-            self._dataframe.columns = tmpcols
-            self._dataframe[colname] = self._dataframe[colname].astype(c_bool)
-            self._dataframe.columns = cols
+            replace_df_column(self._dataframe,colname,self._dataframe[colname].astype(c_bool))
         return
 
     def _fixup_set_value(self):
@@ -687,10 +692,21 @@ def _fixup_set_value(self):
         """
         assert self.data_type == GamsDataType.Set
 
+        warned_values = []
+
+        def fix_individual_value(val):
+            try:
+                return c_bool(val)
+            except:
+                if val not in warned_values:
+                    logger.warn("Replacing {} with {}".format(val,c_bool(True)))
+                    warned_values.append(val)
+                return c_bool(True)
+
         colname = self._dataframe.columns[-1]
         assert colname == self.value_col_names[0], "Unexpected final column in Set dataframe"
         self._dataframe[colname].fillna(value=True,inplace=True)
-        self._dataframe[colname] = self._dataframe[colname].apply(lambda x: c_bool(x))
+        replace_df_column(self._dataframe,colname,self._dataframe[colname].apply(lambda x: c_bool(x)))
         return
 
     @property

setup.py

@@ -3,7 +3,7 @@
 
 setup(
     name = 'gdxpds',
-    version = "1.0.3",
+    version = "1.0.4",
     author = 'Elaine T. Hale',
     author_email = '[email protected]',
     packages = ['gdxpds', 'gdxpds.test'],