robin-sparkless 4.4.0

PySpark-like DataFrame API in Rust on Polars; no JVM.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207

"""
Tests for issue #397: groupBy with aliased columns.

PySpark accepts df.groupBy(F.col('Name').alias('Key')).agg(...).
Sparkless previously raised SparkColumnNotFoundError: cannot resolve 'Key'.
"""

from sparkless.testing import get_imports


class TestIssue397GroupByAlias:
    """Test groupBy with aliased columns (issue #397)."""

    def test_groupby_alias_exact_issue(self, spark):
        """Exact scenario from issue #397."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"Name": "Alice", "Value": 1},
                {"Name": "Alice", "Value": 2},
                {"Name": "Bob", "Value": 3},
                {"Name": "Bob", "Value": 4},
            ]
        )
        df = df.groupBy(F.col("Name").alias("Key")).agg(F.sum("Value"))
        rows = df.collect()
        assert len(rows) == 2
        alice = next(r for r in rows if r["Key"] == "Alice")
        bob = next(r for r in rows if r["Key"] == "Bob")
        assert alice["sum(Value)"] == 3
        assert bob["sum(Value)"] == 7

    def test_groupby_alias_with_show(self, spark):
        """groupBy alias + show() as in issue example."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"Name": "Alice", "Value": 1},
                {"Name": "Alice", "Value": 2},
                {"Name": "Bob", "Value": 3},
            ]
        )
        df = df.groupBy(F.col("Name").alias("Key")).agg(F.sum("Value"))
        df.show()
        rows = df.collect()
        assert len(rows) == 2

    def test_groupby_multiple_cols_one_aliased(self, spark):
        """groupBy with mix of plain and aliased columns."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"A": "x", "B": 1, "Val": 10},
                {"A": "x", "B": 1, "Val": 20},
                {"A": "y", "B": 2, "Val": 30},
            ]
        )
        df = df.groupBy(F.col("A").alias("Key"), "B").agg(F.sum("Val"))
        rows = df.collect()
        assert len(rows) == 2
        key_b_pairs = [(r["Key"], r["B"]) for r in rows]
        assert ("x", 1) in key_b_pairs
        assert ("y", 2) in key_b_pairs

    def test_groupby_all_cols_aliased(self, spark):
        """groupBy with all columns aliased."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"A": 1, "B": 10, "Val": 100},
                {"A": 1, "B": 10, "Val": 200},
                {"A": 2, "B": 20, "Val": 300},
            ]
        )
        df = df.groupBy(F.col("A").alias("X"), F.col("B").alias("Y")).agg(F.sum("Val"))
        rows = df.collect()
        assert len(rows) == 2
        cols = list(rows[0].asDict().keys())
        assert "X" in cols and "Y" in cols
        sums = {(r["X"], r["Y"]): r["sum(Val)"] for r in rows}
        assert sums[(1, 10)] == 300
        assert sums[(2, 20)] == 300

    def test_groupby_alias_avg_max_min(self, spark):
        """groupBy alias with multiple agg functions."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"k": "x", "v": 10},
                {"k": "x", "v": 20},
                {"k": "y", "v": 5},
            ]
        )
        df = df.groupBy(F.col("k").alias("key")).agg(F.avg("v"), F.max("v"), F.min("v"))
        rows = df.collect()
        assert len(rows) == 2
        by_key = {r["key"]: r for r in rows}
        assert by_key["x"]["avg(v)"] == 15.0
        assert by_key["x"]["max(v)"] == 20
        assert by_key["x"]["min(v)"] == 10
        assert by_key["y"]["avg(v)"] == 5.0

    def test_groupby_alias_count(self, spark):
        """groupBy alias with count(*)."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"cat": "A", "x": 1},
                {"cat": "A", "x": 2},
                {"cat": "B", "x": 3},
            ]
        )
        df = df.groupBy(F.col("cat").alias("category")).agg(F.count("*"))
        rows = df.collect()
        assert len(rows) == 2
        # PySpark and sparkless both use count(1) for agg(F.count("*"))
        count_col = "count(1)"
        by_cat = {r["category"]: r[count_col] for r in rows}
        assert by_cat["A"] == 2
        assert by_cat["B"] == 1

    def test_groupby_alias_with_nulls(self, spark):
        """groupBy alias when group column has nulls."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"n": "a", "v": 1},
                {"n": None, "v": 2},
                {"n": "a", "v": 3},
            ]
        )
        df = df.groupBy(F.col("n").alias("name")).agg(F.sum("v"))
        rows = df.collect()
        assert len(rows) == 2
        by_name = {r["name"]: r["sum(v)"] for r in rows}
        assert by_name["a"] == 4
        assert by_name[None] == 2

    def test_groupby_alias_select_after(self, spark):
        """Select after groupBy with alias."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"x": "a", "y": 1},
                {"x": "a", "y": 2},
                {"x": "b", "y": 5},
            ]
        )
        df = (
            df.groupBy(F.col("x").alias("grp"))
            .agg(F.sum("y").alias("total"))
            .select("grp", "total")
        )
        rows = df.collect()
        assert len(rows) == 2
        by_grp = {r["grp"]: r["total"] for r in rows}
        assert by_grp["a"] == 3
        assert by_grp["b"] == 5

    def test_groupby_alias_agg_alias_matches_issue_expected(self, spark):
        """groupBy alias + agg alias to match issue expected output."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"Name": "Alice", "Value": 1},
                {"Name": "Alice", "Value": 2},
                {"Name": "Bob", "Value": 3},
                {"Name": "Bob", "Value": 4},
            ]
        )
        df = df.groupBy(F.col("Name").alias("Key")).agg(
            F.sum("Value").alias("SumValue")
        )
        rows = df.collect()
        assert len(rows) == 2
        assert {r["Key"] for r in rows} == {"Alice", "Bob"}
        by_key = {r["Key"]: r["SumValue"] for r in rows}
        assert by_key["Alice"] == 3
        assert by_key["Bob"] == 7

    def test_groupby_list_syntax_with_alias(self, spark):
        """groupBy([col.alias(...)]) list syntax."""
        imports = get_imports()
        F = imports.F
        df = spark.createDataFrame(
            [
                {"Name": "Alice", "Value": 1},
                {"Name": "Alice", "Value": 2},
                {"Name": "Bob", "Value": 3},
            ]
        )
        df = df.groupBy([F.col("Name").alias("Key")]).agg(F.sum("Value"))
        rows = df.collect()
        assert len(rows) == 2
        assert "Key" in rows[0].asDict()
        by_key = {r["Key"]: r["sum(Value)"] for r in rows}
        assert by_key["Alice"] == 3
        assert by_key["Bob"] == 3