robin-sparkless 4.0.0

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
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362

"""Unit tests for PivotGroupedData aggregate methods (Issue #267).

Uses get_spark_imports() so F and session match the backend (PySpark or Robin).
"""

import pytest

from tests.fixtures.spark_imports import get_spark_imports

imports = get_spark_imports()
F = imports.F


class TestPivotGroupedData:
    """Test PivotGroupedData convenience methods. Uses conftest spark fixture (Robin or PySpark)."""

    @pytest.fixture
    def sample_data(self):
        """Sample data for testing."""
        return [
            {"type": "A", "value": 1},
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]

    def test_pivot_sum(self, spark, sample_data):
        """Test pivot with sum() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).sum("value")
        rows = result.collect()

        assert len(rows) == 2
        # Check column names - should be just pivot values
        schema_names = [f.name for f in result.schema.fields]
        assert "type" in schema_names
        assert "A" in schema_names
        assert "B" in schema_names

        # Find rows by type
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A has values 1 and 10, so sum should be 11
        assert row_a["A"] == 11
        assert row_a["B"] is None  # No B values for type A
        # Type B has value 5
        assert row_b["A"] is None  # No A values for type B
        assert row_b["B"] == 5

    def test_pivot_avg(self, spark, sample_data):
        """Test pivot with avg() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).avg("value")
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: (1 + 10) / 2 = 5.5
        assert row_a["A"] == 5.5
        assert row_a["B"] is None
        # Type B: 5 / 1 = 5.0
        assert row_b["A"] is None
        assert row_b["B"] == 5.0

    def test_pivot_count(self, spark, sample_data):
        """Test pivot with count() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).count()
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A has 2 rows
        assert row_a["A"] == 2
        assert row_a["B"] is None
        # Type B has 1 row
        assert row_b["A"] is None
        assert row_b["B"] == 1

    def test_pivot_max(self, spark, sample_data):
        """Test pivot with max() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).max("value")
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: max(1, 10) = 10
        assert row_a["A"] == 10
        assert row_a["B"] is None
        # Type B: max(5) = 5
        assert row_b["A"] is None
        assert row_b["B"] == 5

    def test_pivot_min(self, spark, sample_data):
        """Test pivot with min() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).min("value")
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: min(1, 10) = 1
        assert row_a["A"] == 1
        assert row_a["B"] is None
        # Type B: min(5) = 5
        assert row_b["A"] is None
        assert row_b["B"] == 5

    def test_pivot_count_distinct(self, spark):
        """Test pivot with count_distinct() method."""
        data = [
            {"type": "A", "value": 1},
            {"type": "A", "value": 1},  # Duplicate
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]
        df = spark.createDataFrame(data)
        result = (
            df.groupBy("type").pivot("type", ["A", "B"]).agg(F.count_distinct("value"))
        )
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: distinct values are 1, 10 = 2
        assert row_a["A"] == 2
        assert row_a["B"] is None
        # Type B: distinct value is 5 = 1
        assert row_b["A"] is None
        assert row_b["B"] == 1

    def test_pivot_collect_list(self, spark):
        """Test pivot with collect_list() method."""
        data = [
            {"type": "A", "value": 1},
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]
        df = spark.createDataFrame(data)
        result = (
            df.groupBy("type").pivot("type", ["A", "B"]).agg(F.collect_list("value"))
        )
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: [1, 10]; no B values (PySpark uses [], Sparkless may use None)
        assert row_a["A"] == [1, 10]
        assert row_a["B"] is None or row_a["B"] == []
        # Type B: [5]; no A values
        assert row_b["A"] is None or row_b["A"] == []
        assert row_b["B"] == [5]

    def test_pivot_collect_set(self, spark):
        """Test pivot with collect_set() method."""
        data = [
            {"type": "A", "value": 1},
            {"type": "A", "value": 1},  # Duplicate
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]
        df = spark.createDataFrame(data)
        result = (
            df.groupBy("type").pivot("type", ["A", "B"]).agg(F.collect_set("value"))
        )
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: {1, 10} (as list); no B values (PySpark uses [], Sparkless may use None)
        assert set(row_a["A"]) == {1, 10}
        assert row_a["B"] is None or row_a["B"] == []
        # Type B: {5}; no A values
        assert row_b["A"] is None or row_b["A"] == []
        assert row_b["B"] == [5]

    def test_pivot_first(self, spark, sample_data):
        """Test pivot with first() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).agg(F.first("value"))
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: first value is 1
        assert row_a["A"] == 1
        assert row_a["B"] is None
        # Type B: first value is 5
        assert row_b["A"] is None
        assert row_b["B"] == 5

    def test_pivot_last(self, spark, sample_data):
        """Test pivot with last() method."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).agg(F.last("value"))
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: last value is 10
        assert row_a["A"] == 10
        assert row_a["B"] is None
        # Type B: last value is 5
        assert row_b["A"] is None
        assert row_b["B"] == 5

    def test_pivot_stddev(self, spark):
        """Test pivot with stddev() method."""
        data = [
            {"type": "A", "value": 1},
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]
        df = spark.createDataFrame(data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).agg(F.stddev("value"))
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: stddev of [1, 10] should be calculated
        assert row_a["A"] is not None
        assert isinstance(row_a["A"], float)
        assert row_a["B"] is None
        # Type B: only one value, stddev should be None
        assert row_b["A"] is None
        assert row_b["B"] is None

    def test_pivot_variance(self, spark):
        """Test pivot with variance() method."""
        data = [
            {"type": "A", "value": 1},
            {"type": "A", "value": 10},
            {"type": "B", "value": 5},
        ]
        df = spark.createDataFrame(data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).agg(F.variance("value"))
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Type A: variance of [1, 10] should be calculated
        assert row_a["A"] is not None
        assert isinstance(row_a["A"], float)
        assert row_a["B"] is None
        # Type B: only one value, variance should be None
        assert row_b["A"] is None
        assert row_b["B"] is None

    def test_pivot_mean(self, spark, sample_data):
        """Test pivot with mean() method (alias for avg)."""
        df = spark.createDataFrame(sample_data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).mean("value")
        rows = result.collect()

        assert len(rows) == 2
        row_a = next((r for r in rows if r["type"] == "A"), None)
        row_b = next((r for r in rows if r["type"] == "B"), None)

        assert row_a is not None
        assert row_b is not None
        # Should be same as avg
        assert row_a["A"] == 5.5
        assert row_a["B"] is None
        assert row_b["A"] is None
        assert row_b["B"] == 5.0

    def test_pivot_multiple_aggregates(self, spark, sample_data):
        """Test pivot with multiple aggregate expressions."""
        df = spark.createDataFrame(sample_data)
        result = (
            df.groupBy("type")
            .pivot("type", ["A", "B"])
            .agg(F.sum("value").alias("total"), F.avg("value").alias("avg_val"))
        )
        rows = result.collect()

        assert len(rows) == 2
        # With multiple expressions, PySpark uses {pivot_value}_{alias} format
        schema_names = [f.name for f in result.schema.fields]
        assert "type" in schema_names
        # Should have columns like A_total, A_avg_val, B_total, B_avg_val
        assert "A_total" in schema_names
        assert "A_avg_val" in schema_names
        assert "B_total" in schema_names
        assert "B_avg_val" in schema_names

    def test_pivot_single_aggregate_with_alias(self, spark, sample_data):
        """Test pivot with single aggregate expression with alias."""
        df = spark.createDataFrame(sample_data)
        result = (
            df.groupBy("type")
            .pivot("type", ["A", "B"])
            .agg(F.sum("value").alias("total"))
        )
        rows = result.collect()

        assert len(rows) == 2
        # PySpark: pivot().agg(one alias) still gives one column per pivot value (A, B).
        # Sparkless may give single column "total". Accept both.
        schema_names = [f.name for f in result.schema.fields]
        assert "type" in schema_names
        assert "total" in schema_names or ("A" in schema_names and "B" in schema_names)

    def test_pivot_empty_groups(self, spark):
        """Test pivot with empty groups."""
        data = [{"type": "A", "value": 1}]
        df = spark.createDataFrame(data)
        result = df.groupBy("type").pivot("type", ["A", "B"]).sum("value")
        rows = result.collect()

        assert len(rows) == 1
        row_a = next((r for r in rows if r["type"] == "A"), None)
        assert row_a is not None
        assert row_a["A"] == 1
        assert row_a["B"] is None  # No B values