work on replcate rows

Author: JohnMount

data_algebra/solutions.py

@@ -482,7 +482,7 @@ def replicate_rows_query(
     seq_column_name: str,
     join_temp_name: str,
     max_count: int,
-) -> Tuple(ViewRepresentation, Any):
+) -> Tuple[ViewRepresentation, Any]:
     """
     Build query to replicate each row by count_column_name copies.
 
@@ -522,7 +522,9 @@ def replicate_rows_query(
     ops = (
         d
             # specify which power table we want to join with
-            .extend({power_key_colname: f'"p" %+% ({count_column_name}.log() / (2).log()).ceil()'})
+            .extend({
+                power_key_colname: f'"p" %+% ({count_column_name}.log() / (2).log()).ceil().as_int64()'
+                })
             # get one row for each number less than or equal to power by under-specified join
             .natural_join(
                 b=TableDescription(

tests/test_cdata_convenience.py

@@ -1,3 +1,4 @@
+
 import re
 import io
 

tests/test_expand_rows.py

@@ -0,0 +1,29 @@
+
+import data_algebra
+import data_algebra.test_util
+from data_algebra.data_ops import descr
+from data_algebra.solutions import replicate_rows_query
+
+
+def test_replicate_rows_query():
+    # get a pandas namespace
+    pd = data_algebra.default_data_model.pd
+    d = pd.DataFrame({
+        'key': ['a', 'b', 'c'],
+        'n': [1, 2, 3],
+    })
+    d_descr = descr(d=d)
+    ops, rt = replicate_rows_query(
+        d_descr,
+        count_column_name='n',
+        seq_column_name='i',
+        join_temp_name='rt',
+        max_count=3,
+        )
+    res = ops.eval({'d': d, 'rt': rt})
+    expect = pd.DataFrame({
+        'key': ['a', 'b', 'b', 'c', 'c', 'c'],
+        'n': [1, 2, 2, 3, 3, 3],
+        'i': ['0', '0', '1', '0', '1', '2'],
+    })
+    assert data_algebra.test_util.equivalent_frames(res, expect)