rebuild and recheck

Author: JohnMount

build/lib/data_algebra/solutions.py

@@ -6,7 +6,7 @@
 
 import numpy
 import data_algebra
-from data_algebra.data_ops import descr, ViewRepresentation
+from data_algebra.data_ops import descr, TableDescription, ViewRepresentation
 from data_algebra.cdata import (
     pivot_specification,
     unpivot_specification,
@@ -473,3 +473,70 @@ def rank_to_average(
         .drop_columns([tie_breaker_column_name])
     )
     return ops
+
+
+def replicate_rows_query(
+    d: ViewRepresentation,
+    *,
+    count_column_name: str,
+    seq_column_name: str,
+    join_temp_name: str,
+    max_count: int,
+) -> Tuple[ViewRepresentation, Any]:
+    """
+    Build query to replicate each row by count_column_name copies.
+
+    :param d: incoming data description.
+    :param count_column_name: name of count column, should be non-negative integers.
+    :param seq_column_name: name of colulmn to land sequence in.
+    :param join_temp_name: name for join temp table.
+    :param max_count: maximum in count column we need to handle, should be a reasonable upper bound.
+    :return: ops and table to join against
+    """
+    assert isinstance(d, TableDescription)
+    assert isinstance(count_column_name, str)
+    assert count_column_name in d.column_names
+    assert isinstance(seq_column_name, str)
+    assert seq_column_name not in d.column_names
+    assert isinstance(join_temp_name, str)
+    assert isinstance(max_count, int)
+    assert max_count > 0
+    # reserve a power key column
+    power_key_colname = 'power'
+    assert power_key_colname != count_column_name
+    assert power_key_colname not in d.column_names
+    # get a pandas namespace
+    pd = data_algebra.default_data_model.pd
+    # build powers of 2 until max_count is met or exceeded
+    powers = list(range(int(numpy.ceil(numpy.log(max_count)/numpy.log(2))) + 1))
+    # replicate each power the number of times it specifies
+    count_frame = pd.concat([
+            pd.DataFrame({
+                power_key_colname: f'p{p}',
+                seq_column_name: range(int(2 ** p)),
+            })
+            for p in powers
+        ])
+    count_frame.reset_index(drop=True, inplace=True)
+    # specify ops that produce row replicates
+    ops = (
+        d
+            # specify which power table we want to join with
+            .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(
+                    table_name=join_temp_name,
+                    column_names=[power_key_colname, seq_column_name],
+                    ),
+                by=[power_key_colname],
+                jointype='inner',
+                )
+            # drop rows exceeding desired count
+            .select_rows(f'{seq_column_name} < {count_column_name}')
+            # drop the power group column id
+            .drop_columns([power_key_colname])
+    )
+    return ops, count_frame

coverage.txt

@@ -2,7 +2,7 @@
 platform darwin -- Python 3.9.12, pytest-7.1.1, pluggy-1.0.0
 rootdir: /Users/johnmount/Documents/work/data_algebra
 plugins: anyio-3.5.0, cov-3.0.0
-collected 302 items
+collected 303 items
 
 tests/test_OrderedSet.py ......                                          [  1%]
 tests/test_agg.py .                                                      [  2%]
@@ -36,13 +36,14 @@ tests/test_ex_examples.py ............                                   [ 24%]
 tests/test_example1.py ....                                              [ 26%]
 tests/test_exp.py .                                                      [ 26%]
 tests/test_exp_parens.py .                                               [ 26%]
+tests/test_expand_rows.py .                                              [ 27%]
 tests/test_expr_parse.py .                                               [ 27%]
-tests/test_expression_expectations.py ..                                 [ 27%]
-tests/test_extend.py ......                                              [ 29%]
+tests/test_expression_expectations.py ..                                 [ 28%]
+tests/test_extend.py ......                                              [ 30%]
 tests/test_extend_order.py .                                             [ 30%]
 tests/test_first_last.py .                                               [ 30%]
 tests/test_float_divide.py ...                                           [ 31%]
-tests/test_flow_text.py .                                                [ 31%]
+tests/test_flow_text.py .                                                [ 32%]
 tests/test_forbidden_calculation.py ..                                   [ 32%]
 tests/test_forbidden_ops.py ...                                          [ 33%]
 tests/test_free_fn.py .                                                  [ 33%]
@@ -71,15 +72,15 @@ tests/test_natural_join.py .                                             [ 50%]
 tests/test_neg.py .                                                      [ 50%]
 tests/test_null_bad.py ..                                                [ 51%]
 tests/test_one_row_cdata_convert.py .                                    [ 51%]
-tests/test_ops.py .                                                      [ 51%]
+tests/test_ops.py .                                                      [ 52%]
 tests/test_ops_eq.py .                                                   [ 52%]
-tests/test_or.py .....                                                   [ 53%]
+tests/test_or.py .....                                                   [ 54%]
 tests/test_order_limit.py .                                              [ 54%]
 tests/test_ordered_agg_group.py .                                        [ 54%]
 tests/test_parens.py ..                                                  [ 55%]
-tests/test_parse.py ..........................                           [ 63%]
+tests/test_parse.py ..........................                           [ 64%]
 tests/test_project.py .....                                              [ 65%]
-tests/test_rank.py .                                                     [ 65%]
+tests/test_rank.py .                                                     [ 66%]
 tests/test_rank_to_average.py .                                          [ 66%]
 tests/test_ranked_example.py .                                           [ 66%]
 tests/test_readme_example.py .                                           [ 66%]
@@ -150,12 +151,12 @@ data_algebra/pandas_base.py              501     44    91%   29, 69, 71, 84, 94,
 data_algebra/pandas_model.py               5      0   100%
 data_algebra/parse_by_lark.py            164     24    85%   71, 93, 108, 129-130, 137, 161, 171, 185-186, 188, 200, 206, 213-217, 245, 253, 263-266
 data_algebra/python3_lark.py               1      0   100%
-data_algebra/solutions.py                117      4    97%   63, 290, 371, 454
+data_algebra/solutions.py                135      4    97%   63, 290, 371, 454
 data_algebra/sql_format_options.py        15      2    87%   61, 69
 data_algebra/test_util.py                290     56    81%   88, 110, 120, 123, 127, 150, 153, 157, 159-162, 173, 228-229, 245-250, 254, 266, 268-276, 322, 325, 333, 344, 351, 355, 367, 380, 463-466, 468-471, 473-476, 478-481, 555-560, 566-567, 570
 data_algebra/util.py                     140     29    79%   26, 51, 56, 61, 84-85, 88-89, 92-93, 96-97, 100-101, 104-105, 108-109, 112-113, 116-117, 120-121, 192, 207, 248, 252, 254
 --------------------------------------------------------------------
-TOTAL                                   5759    845    85%
+TOTAL                                   5777    845    85%
 
 
-======================= 302 passed in 417.20s (0:06:57) ========================
+======================= 303 passed in 423.88s (0:07:03) ========================