kumoai 2.14.0.dev202512181731__cp312-cp312-macosx_11_0_arm64.whl → 2.14.0.dev202601041732__cp312-cp312-macosx_11_0_arm64.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- kumoai/__init__.py +23 -26
- kumoai/_version.py +1 -1
- kumoai/client/client.py +6 -0
- kumoai/client/jobs.py +24 -0
- kumoai/connector/utils.py +21 -7
- kumoai/experimental/rfm/__init__.py +24 -22
- kumoai/experimental/rfm/backend/local/graph_store.py +12 -21
- kumoai/experimental/rfm/backend/local/sampler.py +0 -3
- kumoai/experimental/rfm/backend/local/table.py +25 -24
- kumoai/experimental/rfm/backend/snow/sampler.py +106 -61
- kumoai/experimental/rfm/backend/snow/table.py +146 -51
- kumoai/experimental/rfm/backend/sqlite/sampler.py +127 -78
- kumoai/experimental/rfm/backend/sqlite/table.py +94 -47
- kumoai/experimental/rfm/base/__init__.py +6 -7
- kumoai/experimental/rfm/base/column.py +97 -5
- kumoai/experimental/rfm/base/expression.py +44 -0
- kumoai/experimental/rfm/base/sampler.py +5 -17
- kumoai/experimental/rfm/base/source.py +1 -1
- kumoai/experimental/rfm/base/sql_sampler.py +68 -9
- kumoai/experimental/rfm/base/table.py +291 -126
- kumoai/experimental/rfm/graph.py +139 -86
- kumoai/experimental/rfm/infer/__init__.py +6 -4
- kumoai/experimental/rfm/infer/dtype.py +6 -1
- kumoai/experimental/rfm/infer/multicategorical.py +1 -1
- kumoai/experimental/rfm/infer/stype.py +35 -0
- kumoai/experimental/rfm/relbench.py +76 -0
- kumoai/experimental/rfm/rfm.py +30 -42
- kumoai/experimental/rfm/task_table.py +247 -0
- kumoai/trainer/distilled_trainer.py +175 -0
- kumoai/utils/display.py +51 -0
- {kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/METADATA +1 -1
- {kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/RECORD +35 -31
- kumoai/experimental/rfm/base/column_expression.py +0 -16
- kumoai/experimental/rfm/base/sql_table.py +0 -113
- {kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/WHEEL +0 -0
- {kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/licenses/LICENSE +0 -0
- {kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/top_level.txt +0 -0
|
@@ -1,113 +0,0 @@
|
|
|
1
|
-
from abc import abstractmethod
|
|
2
|
-
from collections import defaultdict
|
|
3
|
-
from collections.abc import Sequence
|
|
4
|
-
from functools import cached_property
|
|
5
|
-
from typing import Any
|
|
6
|
-
|
|
7
|
-
from kumoapi.model_plan import MissingType
|
|
8
|
-
|
|
9
|
-
from kumoai.experimental.rfm.base import (
|
|
10
|
-
ColumnExpressionType,
|
|
11
|
-
SourceForeignKey,
|
|
12
|
-
Table,
|
|
13
|
-
)
|
|
14
|
-
from kumoai.utils import quote_ident
|
|
15
|
-
|
|
16
|
-
|
|
17
|
-
class SQLTable(Table):
|
|
18
|
-
r"""A :class:`SQLTable` specifies a :class:`Table` backed by a SQL
|
|
19
|
-
database.
|
|
20
|
-
|
|
21
|
-
Args:
|
|
22
|
-
name: The logical name of this table.
|
|
23
|
-
source_name: The physical name of this table in the database. If set to
|
|
24
|
-
``None``, ``name`` is being used.
|
|
25
|
-
columns: The selected physical columns of this table.
|
|
26
|
-
column_expressions: The logical columns of this table.
|
|
27
|
-
primary_key: The name of the primary key of this table, if it exists.
|
|
28
|
-
time_column: The name of the time column of this table, if it exists.
|
|
29
|
-
end_time_column: The name of the end time column of this table, if it
|
|
30
|
-
exists.
|
|
31
|
-
"""
|
|
32
|
-
def __init__(
|
|
33
|
-
self,
|
|
34
|
-
name: str,
|
|
35
|
-
source_name: str | None = None,
|
|
36
|
-
columns: Sequence[str] | None = None,
|
|
37
|
-
column_expressions: Sequence[ColumnExpressionType] | None = None,
|
|
38
|
-
primary_key: MissingType | str | None = MissingType.VALUE,
|
|
39
|
-
time_column: str | None = None,
|
|
40
|
-
end_time_column: str | None = None,
|
|
41
|
-
) -> None:
|
|
42
|
-
|
|
43
|
-
self._connection: Any
|
|
44
|
-
self._source_name = source_name or name
|
|
45
|
-
|
|
46
|
-
super().__init__(
|
|
47
|
-
name=name,
|
|
48
|
-
columns=[],
|
|
49
|
-
primary_key=None,
|
|
50
|
-
time_column=None,
|
|
51
|
-
end_time_column=None,
|
|
52
|
-
)
|
|
53
|
-
|
|
54
|
-
if isinstance(primary_key, MissingType):
|
|
55
|
-
primary_key = self._source_primary_key
|
|
56
|
-
|
|
57
|
-
# Add column expressions with highest priority:
|
|
58
|
-
self._add_column_expressions(column_expressions or [])
|
|
59
|
-
|
|
60
|
-
if columns is None:
|
|
61
|
-
for column_name in self._source_column_dict.keys():
|
|
62
|
-
if column_name not in self:
|
|
63
|
-
self.add_column(column_name)
|
|
64
|
-
else:
|
|
65
|
-
for column_name in columns:
|
|
66
|
-
self.add_column(column_name)
|
|
67
|
-
|
|
68
|
-
if primary_key is not None:
|
|
69
|
-
if primary_key not in self:
|
|
70
|
-
self.add_column(primary_key)
|
|
71
|
-
self.primary_key = primary_key
|
|
72
|
-
|
|
73
|
-
if time_column is not None:
|
|
74
|
-
if time_column not in self:
|
|
75
|
-
self.add_column(time_column)
|
|
76
|
-
self.time_column = time_column
|
|
77
|
-
|
|
78
|
-
if end_time_column is not None:
|
|
79
|
-
if end_time_column not in self:
|
|
80
|
-
self.add_column(end_time_column)
|
|
81
|
-
self.end_time_column = end_time_column
|
|
82
|
-
|
|
83
|
-
@property
|
|
84
|
-
def fqn(self) -> str:
|
|
85
|
-
r"""The fully-qualified quoted source table name."""
|
|
86
|
-
return quote_ident(self._source_name)
|
|
87
|
-
|
|
88
|
-
# Column ##################################################################
|
|
89
|
-
|
|
90
|
-
def _add_column_expressions(
|
|
91
|
-
self,
|
|
92
|
-
columns: Sequence[ColumnExpressionType],
|
|
93
|
-
) -> None:
|
|
94
|
-
pass
|
|
95
|
-
|
|
96
|
-
# Abstract Methods ########################################################
|
|
97
|
-
|
|
98
|
-
@cached_property
|
|
99
|
-
def _source_foreign_key_dict(self) -> dict[str, SourceForeignKey]:
|
|
100
|
-
fkeys = self._get_source_foreign_keys()
|
|
101
|
-
# NOTE Drop all keys that link to multiple keys in the same table since
|
|
102
|
-
# we don't support composite keys yet:
|
|
103
|
-
table_pkeys: dict[str, set[str]] = defaultdict(set)
|
|
104
|
-
for fkey in fkeys:
|
|
105
|
-
table_pkeys[fkey.dst_table].add(fkey.primary_key)
|
|
106
|
-
return {
|
|
107
|
-
fkey.name: fkey
|
|
108
|
-
for fkey in fkeys if len(table_pkeys[fkey.dst_table]) == 1
|
|
109
|
-
}
|
|
110
|
-
|
|
111
|
-
@abstractmethod
|
|
112
|
-
def _get_source_foreign_keys(self) -> list[SourceForeignKey]:
|
|
113
|
-
pass
|
|
File without changes
|
{kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/licenses/LICENSE
RENAMED
|
File without changes
|
{kumoai-2.14.0.dev202512181731.dist-info → kumoai-2.14.0.dev202601041732.dist-info}/top_level.txt
RENAMED
|
File without changes
|