work on windowed eval

Author: JohnMount

build/lib/data_algebra/data_ops.py

@@ -867,7 +867,7 @@ def __init__(
         reverse=None,
         parse_env=None
     ):
-        partitioned = False
+        windowed_situation = False
         if ops is None:
             ops = {}
         ops = data_algebra.expr_rep.parse_assignments_in_context(
@@ -878,25 +878,25 @@ def __init__(
         for (k, opk) in ops.items():  # look for aggregation functions
             if isinstance(opk, data_algebra.expr_rep.Expression):
                 if opk.op in data_algebra.expr_rep.fn_names_that_imply_windowed_situation:
-                    partitioned = True
+                    windowed_situation = True
         self.ops = ops
         if partition_by is None:
             partition_by = []
         if isinstance(partition_by, numbers.Number):
             partition_by = []
-            partitioned = True
+            windowed_situation = True
         if isinstance(partition_by, str):
             partition_by = [partition_by]
         if len(partition_by) > 0:
-            partitioned = True
+            windowed_situation = True
         self.partition_by = partition_by
         if order_by is None:
             order_by = []
         if isinstance(order_by, str):
             order_by = [order_by]
         if len(order_by) > 0:
-            partitioned = True
-        self.partitioned = partitioned
+            windowed_situation = True
+        self.windowed_situation = windowed_situation
         self.order_by = order_by
         if reverse is None:
             reverse = []
@@ -935,20 +935,20 @@ def __init__(
         if len(bad_overwrite) > 0:
             raise ValueError("tried to change: " + str(bad_overwrite))
         # check op arguments are very simple: all arguments are column names
-        if partitioned:
+        if windowed_situation:
             for (k, opk) in ops.items():
                 if not isinstance(opk, data_algebra.expr_rep.Expression):
                     raise ValueError(
                         "non-aggregated expression in windowed/partitoned extend: " +
                         "'" + k + "': '" + opk.to_pandas() + "'"
                     )
                 if len(opk.args) > 1:
-                    raise ValueError("in windowed/partitioned situations only simple operators are allowed, " +
+                    raise ValueError("in windowed situations only simple operators are allowed, " +
                                      "'" + k + "': '" + opk.to_pandas() + "' term is too complex an expression")
                 if len(opk.args) > 0:
                     value_name = opk.args[0].to_pandas()
                     if value_name not in source.column_set:
-                        raise ValueError("in windowed/partitioned situations only simple operators are allowed, " +
+                        raise ValueError("in windowed situations only simple operators are allowed, " +
                                          "'" + k + "': '" + opk.to_pandas() + "' term is too complex an expression")
         ViewRepresentation.__init__(self, column_names=column_names, sources=[source])
 
@@ -994,7 +994,7 @@ def to_python_implementation(self, *, indent=0, strict=True, print_sources=True)
         ]
         flowed = ("," + spacer + " ").join(ops)
         s = s + ("extend({" + spacer + " " + flowed + "}")
-        if self.partitioned:
+        if self.windowed_situation:
             if len(self.partition_by) > 0:
                 s = s + "," + spacer + "partition_by=" + self.partition_by.__repr__()
             else:

build/lib/data_algebra/db_model.py

@@ -243,7 +243,7 @@ def extend_to_sql(self, extend_node, *, using=None, temp_id_source=None):
         sub_view_name = "SQ_" + str(temp_id_source[0])
         temp_id_source[0] = temp_id_source[0] + 1
         window_term = ""
-        if extend_node.partitioned or (len(extend_node.partition_by) > 0) or (len(extend_node.order_by) > 0):
+        if extend_node.windowed_situation or (len(extend_node.partition_by) > 0) or (len(extend_node.order_by) > 0):
             window_term = " OVER ( "
             if len(extend_node.partition_by) > 0:
                 pt = [self.quote_identifier(ci) for ci in extend_node.partition_by]

build/lib/data_algebra/pandas_model.py

@@ -67,7 +67,7 @@ def table_step(self, op, *, data_map, eval_env):
     def extend_step(self, op, *, data_map, eval_env):
         if not isinstance(op, data_algebra.data_ops.ExtendNode):
             raise TypeError("op was supposed to be a data_algebra.data_ops.ExtendNode")
-        window_situation = op.partitioned or (len(op.partition_by) > 0) or (len(op.order_by) > 0)
+        window_situation = op.windowed_situation or (len(op.partition_by) > 0) or (len(op.order_by) > 0)
         if window_situation:
             self.check_extend_window_fns(op)
         res = op.sources[0].eval_implementation(

coverage.txt

@@ -71,4 +71,4 @@ data_algebra/yaml.py                     123     15    88%
 TOTAL                                   3404    821    76%
 
 
-============================== 60 passed in 6.27s ==============================
+============================== 60 passed in 6.29s ==============================

data_algebra/data_ops.py

@@ -867,7 +867,7 @@ def __init__(
         reverse=None,
         parse_env=None
     ):
-        partitioned = False
+        windowed_situation = False
         if ops is None:
             ops = {}
         ops = data_algebra.expr_rep.parse_assignments_in_context(
@@ -878,25 +878,25 @@ def __init__(
         for (k, opk) in ops.items():  # look for aggregation functions
             if isinstance(opk, data_algebra.expr_rep.Expression):
                 if opk.op in data_algebra.expr_rep.fn_names_that_imply_windowed_situation:
-                    partitioned = True
+                    windowed_situation = True
         self.ops = ops
         if partition_by is None:
             partition_by = []
         if isinstance(partition_by, numbers.Number):
             partition_by = []
-            partitioned = True
+            windowed_situation = True
         if isinstance(partition_by, str):
             partition_by = [partition_by]
         if len(partition_by) > 0:
-            partitioned = True
+            windowed_situation = True
         self.partition_by = partition_by
         if order_by is None:
             order_by = []
         if isinstance(order_by, str):
             order_by = [order_by]
         if len(order_by) > 0:
-            partitioned = True
-        self.partitioned = partitioned
+            windowed_situation = True
+        self.windowed_situation = windowed_situation
         self.order_by = order_by
         if reverse is None:
             reverse = []
@@ -935,20 +935,20 @@ def __init__(
         if len(bad_overwrite) > 0:
             raise ValueError("tried to change: " + str(bad_overwrite))
         # check op arguments are very simple: all arguments are column names
-        if partitioned:
+        if windowed_situation:
             for (k, opk) in ops.items():
                 if not isinstance(opk, data_algebra.expr_rep.Expression):
                     raise ValueError(
                         "non-aggregated expression in windowed/partitoned extend: " +
                         "'" + k + "': '" + opk.to_pandas() + "'"
                     )
                 if len(opk.args) > 1:
-                    raise ValueError("in windowed/partitioned situations only simple operators are allowed, " +
+                    raise ValueError("in windowed situations only simple operators are allowed, " +
                                      "'" + k + "': '" + opk.to_pandas() + "' term is too complex an expression")
                 if len(opk.args) > 0:
                     value_name = opk.args[0].to_pandas()
                     if value_name not in source.column_set:
-                        raise ValueError("in windowed/partitioned situations only simple operators are allowed, " +
+                        raise ValueError("in windowed situations only simple operators are allowed, " +
                                          "'" + k + "': '" + opk.to_pandas() + "' term is too complex an expression")
         ViewRepresentation.__init__(self, column_names=column_names, sources=[source])
 
@@ -994,7 +994,7 @@ def to_python_implementation(self, *, indent=0, strict=True, print_sources=True)
         ]
         flowed = ("," + spacer + " ").join(ops)
         s = s + ("extend({" + spacer + " " + flowed + "}")
-        if self.partitioned:
+        if self.windowed_situation:
             if len(self.partition_by) > 0:
                 s = s + "," + spacer + "partition_by=" + self.partition_by.__repr__()
             else:

data_algebra/db_model.py

@@ -243,7 +243,7 @@ def extend_to_sql(self, extend_node, *, using=None, temp_id_source=None):
         sub_view_name = "SQ_" + str(temp_id_source[0])
         temp_id_source[0] = temp_id_source[0] + 1
         window_term = ""
-        if extend_node.partitioned or (len(extend_node.partition_by) > 0) or (len(extend_node.order_by) > 0):
+        if extend_node.windowed_situation or (len(extend_node.partition_by) > 0) or (len(extend_node.order_by) > 0):
             window_term = " OVER ( "
             if len(extend_node.partition_by) > 0:
                 pt = [self.quote_identifier(ci) for ci in extend_node.partition_by]

data_algebra/pandas_model.py

@@ -67,7 +67,7 @@ def table_step(self, op, *, data_map, eval_env):
     def extend_step(self, op, *, data_map, eval_env):
         if not isinstance(op, data_algebra.data_ops.ExtendNode):
             raise TypeError("op was supposed to be a data_algebra.data_ops.ExtendNode")
-        window_situation = op.partitioned or (len(op.partition_by) > 0) or (len(op.order_by) > 0)
+        window_situation = op.windowed_situation or (len(op.partition_by) > 0) or (len(op.order_by) > 0)
         if window_situation:
             self.check_extend_window_fns(op)
         res = op.sources[0].eval_implementation(

dist/data_algebra-0.3.4-py3-none-any.whl

@@ -107,10 +107,10 @@ def test_example1_1_detect_agg():
         extend({
         'y_mean': 'y.mean()'
         })
-    assert ops.partitioned
+    assert ops.windowed_situation
 
     ops2 = table_desc. \
         extend({
         'y_mean': 'y+1'
         })
-    assert not ops2.partitioned
+    assert not ops2.windowed_situation