Updated info and memory_usage to new backend

modin/data_management/data_manager.py

@@ -522,6 +522,27 @@ def idxmin_builder(df, **kwargs):
         # have to do a conversion.
         return self._post_process_idx_ops(axis, min_result)
 
+    def info(self, **kwargs):
+        def info_builder(df, **kwargs):
+            result = pandas.DataFrame()
+            if memory_usage:
+                result['memory'] = df.memory_usage(index=False, deep=memory_usage_deep)
+            if null_counts:
+                result['count'] = df.count(axis=0)
+            return result
+
+        memory_usage = kwargs.get('memory_usage', True)
+        null_counts = kwargs.get('null_counts', True)
+
+        if type(memory_usage) == str and memory_usage == 'deep':
+            memory_usage_deep = True
+        else:
+            memory_usage_deep = False
+
+        func = self._prepare_method(info_builder, **kwargs)
+        return self.full_axis_reduce(func, 0)
+
+
     def first_valid_index(self):
 
         # It may be possible to incrementally check each partition, but this
@@ -556,6 +577,14 @@ def median(self, **kwargs):
         func = self._prepare_method(pandas.DataFrame.median, **kwargs)
         return self.full_axis_reduce(func, axis)
 
+    def memory_usage(self, **kwargs):
+        def memory_usage_builder(df, **kwargs):
+            return df.memory_usage(index=False, deep=deep)
+
+        deep = kwargs.get('deep', False)
+        func = self._prepare_method(memory_usage_builder, **kwargs)
+        return self.full_axis_reduce(func, 0)
+
     def nunique(self, **kwargs):
         axis = kwargs.get("axis", 0)
         func = self._prepare_method(pandas.DataFrame.nunique, **kwargs)
@@ -599,7 +628,7 @@ def query(self, expr, **kwargs):
         cls = type(self)
         columns = self.columns
 
-        def query_builder(df):
+        def query_builder(df, **kwargs):
             # This is required because of an Arrow limitation
             # TODO revisit for Arrow error
             df = df.copy()
@@ -616,6 +645,33 @@ def query_builder(df):
 
         return cls(new_data, new_index, self.columns)
 
+    def eval(self, expr, **kwargs):
+        cls = type(self)
+        columns = self.columns
+
+        def eval_builder(df, **kwargs):
+            df.columns = columns
+            result = df.eval(expr, inplace=False, **kwargs)
+            # If result is a series, expr was not an assignment expression.
+            if not isinstance(result, pandas.Series):
+                result.columns = pandas.RangeIndex(0, len(result.columns))
+            return result
+
+        func = self._prepare_method(eval_builder, **kwargs)
+        new_data = self.map_across_full_axis(1, func)
+
+        # eval can update the columns, so we must update columns
+        columns_copy = pandas.DataFrame(columns=columns)
+        columns_copy = columns_copy.eval(expr, inplace=False, **kwargs)
+        if isinstance(columns_copy, pandas.Series):
+            # To create a data manager, we need the 
+            # columns to be in a list-like
+            columns = list(columns_copy.name)
+        else:
+            columns = columns_copy.columns
+
+        return cls(new_data, self.index, columns)
+
     def quantile_for_list_of_values(self, **kwargs):
         cls = type(self)
         axis = kwargs.get("axis", 0)