Implement chain group_by (#482)

* Implement chain group_by
* Use 'sql_to_python' for GenericFunction type conversion

src/datachain/__init__.py

@@ -1,3 +1,4 @@
+from datachain.lib import func
 from datachain.lib.data_model import DataModel, DataType, is_chain_type
 from datachain.lib.dc import C, Column, DataChain, Sys
 from datachain.lib.file import (
@@ -34,6 +35,7 @@
     "Sys",
     "TarVFile",
     "TextFile",
+    "func",
     "is_chain_type",
     "metrics",
     "param",

src/datachain/data_storage/sqlite.py

@@ -763,6 +763,14 @@ def copy_table(
         query: Select,
         progress_cb: Optional[Callable[[int], None]] = None,
     ) -> None:
+        if len(query._group_by_clause) > 0:
+            select_q = query.with_only_columns(
+                *[c for c in query.selected_columns if c.name != "sys__id"]
+            )
+            q = table.insert().from_select(list(select_q.selected_columns), select_q)
+            self.db.execute(q)
+            return
+
         if "sys__id" in query.selected_columns:
             col_id = query.selected_columns.sys__id
         else:

src/datachain/lib/convert/sql_to_python.py

@@ -4,15 +4,11 @@
 from sqlalchemy import ColumnElement
 
 
-def sql_to_python(args_map: dict[str, ColumnElement]) -> dict[str, Any]:
-    res = {}
-    for name, sql_exp in args_map.items():
-        try:
-            type_ = sql_exp.type.python_type
-            if type_ == Decimal:
-                type_ = float
-        except NotImplementedError:
-            type_ = str
-        res[name] = type_
-
-    return res
+def sql_to_python(sql_exp: ColumnElement) -> Any:
+    try:
+        type_ = sql_exp.type.python_type
+        if type_ == Decimal:
+            type_ = float
+    except NotImplementedError:
+        type_ = str
+    return type_

src/datachain/lib/dc.py

@@ -29,6 +29,7 @@
 from datachain.lib.dataset_info import DatasetInfo
 from datachain.lib.file import ArrowRow, File, get_file_type
 from datachain.lib.file import ExportPlacement as FileExportPlacement
+from datachain.lib.func import Func
 from datachain.lib.listing import (
     is_listing_dataset,
     is_listing_expired,
@@ -42,21 +43,12 @@
 from datachain.lib.model_store import ModelStore
 from datachain.lib.settings import Settings
 from datachain.lib.signal_schema import SignalSchema
-from datachain.lib.udf import (
-    Aggregator,
-    BatchMapper,
-    Generator,
-    Mapper,
-    UDFBase,
-)
+from datachain.lib.udf import Aggregator, BatchMapper, Generator, Mapper, UDFBase
 from datachain.lib.udf_signature import UdfSignature
-from datachain.lib.utils import DataChainParamsError
+from datachain.lib.utils import DataChainColumnError, DataChainParamsError
 from datachain.query import Session
-from datachain.query.dataset import (
-    DatasetQuery,
-    PartitionByType,
-)
-from datachain.query.schema import DEFAULT_DELIMITER, Column
+from datachain.query.dataset import DatasetQuery, PartitionByType
+from datachain.query.schema import DEFAULT_DELIMITER, Column, ColumnMeta
 from datachain.sql.functions import path as pathfunc
 from datachain.telemetry import telemetry
 from datachain.utils import batched_it, inside_notebook
@@ -149,11 +141,6 @@ def _get_str(on: Sequence[Union[str, sqlalchemy.ColumnElement]]) -> str:
         super().__init__(f"Merge error on='{on_str}'{right_on_str}: {msg}")
 
 
-class DataChainColumnError(DataChainParamsError):  # noqa: D101
-    def __init__(self, col_name, msg):  # noqa: D107
-        super().__init__(f"Error for column {col_name}: {msg}")
-
-
 OutputType = Union[None, DataType, Sequence[str], dict[str, DataType]]
 
 
@@ -982,10 +969,9 @@ def distinct(self, arg: str, *args: str) -> "Self":  # type: ignore[override]
         row is left in the result set.
 
         Example:
-        ```py
-         dc.distinct("file.parent", "file.name")
-        )
-        ```
+            ```py
+            dc.distinct("file.parent", "file.name")
+            ```
         """
         return self._evolve(
             query=self._query.distinct(
@@ -1011,6 +997,60 @@ def select_except(self, *args: str) -> "Self":
             query=self._query.select(*columns), signal_schema=new_schema
         )
 
+    def group_by(
+        self,
+        *,
+        partition_by: Union[str, Sequence[str]],
+        **kwargs: Func,
+    ) -> "Self":
+        """Group rows by specified set of signals and return new signals
+        with aggregated values.
+
+        Example:
+            ```py
+            chain = chain.group_by(
+                cnt=func.count(),
+                partition_by=("file_source", "file_ext"),
+            )
+            ```
+        """
+        if isinstance(partition_by, str):
+            partition_by = [partition_by]
+        if not partition_by:
+            raise ValueError("At least one column should be provided for partition_by")
+
+        if not kwargs:
+            raise ValueError("At least one column should be provided for group_by")
+        for col_name, func in kwargs.items():
+            if not isinstance(func, Func):
+                raise DataChainColumnError(
+                    col_name,
+                    f"Column {col_name} has type {type(func)} but expected Func object",
+                )
+
+        partition_by_columns: list[Column] = []
+        signal_columns: list[Column] = []
+        schema_fields: dict[str, DataType] = {}
+
+        # validate partition_by columns and add them to the schema
+        for col_name in partition_by:
+            col_db_name = ColumnMeta.to_db_name(col_name)
+            col_type = self.signals_schema.get_column_type(col_db_name)
+            col = Column(col_db_name, python_to_sql(col_type))
+            partition_by_columns.append(col)
+            schema_fields[col_db_name] = col_type
+
+        # validate signal columns and add them to the schema
+        for col_name, func in kwargs.items():
+            col = func.get_column(self.signals_schema, label=col_name)
+            signal_columns.append(col)
+            schema_fields[col_name] = func.get_result_type(self.signals_schema)
+
+        return self._evolve(
+            query=self._query.group_by(signal_columns, partition_by_columns),
+            signal_schema=SignalSchema(schema_fields),
+        )
+
     def mutate(self, **kwargs) -> "Self":
         """Create new signals based on existing signals.
 

src/datachain/lib/func/__init__.py

@@ -0,0 +1,14 @@
+from .aggregate import any_value, avg, collect, concat, count, max, min, sum
+from .func import Func
+
+__all__ = [
+    "Func",
+    "any_value",
+    "avg",
+    "collect",
+    "concat",
+    "count",
+    "max",
+    "min",
+    "sum",
+]

src/datachain/lib/func/aggregate.py

@@ -0,0 +1,42 @@
+from typing import Optional
+
+from sqlalchemy import func as sa_func
+
+from datachain.sql import functions as dc_func
+
+from .func import Func
+
+
+def count(col: Optional[str] = None) -> Func:
+    return Func(inner=sa_func.count, col=col, result_type=int)
+
+
+def sum(col: str) -> Func:
+    return Func(inner=sa_func.sum, col=col)
+
+
+def avg(col: str) -> Func:
+    return Func(inner=dc_func.aggregate.avg, col=col)
+
+
+def min(col: str) -> Func:
+    return Func(inner=sa_func.min, col=col)
+
+
+def max(col: str) -> Func:
+    return Func(inner=sa_func.max, col=col)
+
+
+def any_value(col: str) -> Func:
+    return Func(inner=dc_func.aggregate.any_value, col=col)
+
+
+def collect(col: str) -> Func:
+    return Func(inner=dc_func.aggregate.collect, col=col, is_array=True)
+
+
+def concat(col: str, separator="") -> Func:
+    def inner(arg):
+        return dc_func.aggregate.group_concat(arg, separator)
+
+    return Func(inner=inner, col=col, result_type=str)

src/datachain/lib/func/func.py

@@ -0,0 +1,64 @@
+from typing import TYPE_CHECKING, Callable, Optional
+
+from datachain.lib.convert.python_to_sql import python_to_sql
+from datachain.lib.utils import DataChainColumnError
+from datachain.query.schema import Column, ColumnMeta
+
+if TYPE_CHECKING:
+    from datachain import DataType
+    from datachain.lib.signal_schema import SignalSchema
+
+
+class Func:
+    def __init__(
+        self,
+        inner: Callable,
+        col: Optional[str] = None,
+        result_type: Optional["DataType"] = None,
+        is_array: bool = False,
+    ) -> None:
+        self.inner = inner
+        self.col = col
+        self.result_type = result_type
+        self.is_array = is_array
+
+    @property
+    def db_col(self) -> Optional[str]:
+        return ColumnMeta.to_db_name(self.col) if self.col else None
+
+    def db_col_type(self, signals_schema: "SignalSchema") -> Optional["DataType"]:
+        if not self.db_col:
+            return None
+        col_type: type = signals_schema.get_column_type(self.db_col)
+        return list[col_type] if self.is_array else col_type  # type: ignore[valid-type]
+
+    def get_result_type(self, signals_schema: "SignalSchema") -> "DataType":
+        col_type = self.db_col_type(signals_schema)
+
+        if self.result_type:
+            return self.result_type
+
+        if col_type:
+            return col_type
+
+        raise DataChainColumnError(
+            str(self.inner),
+            "Column name is required to infer result type",
+        )
+
+    def get_column(
+        self, signals_schema: "SignalSchema", label: Optional[str] = None
+    ) -> Column:
+        if self.col:
+            if label == "collect":
+                print(label)
+            col_type = self.get_result_type(signals_schema)
+            col = Column(self.db_col, python_to_sql(col_type))
+            func_col = self.inner(col)
+        else:
+            func_col = self.inner()
+
+        if label:
+            func_col = func_col.label(label)
+
+        return func_col

src/datachain/lib/signal_schema.py

@@ -400,6 +400,12 @@ def _set_file_stream(
             if ModelStore.is_pydantic(finfo.annotation):
                 SignalSchema._set_file_stream(getattr(obj, field), catalog, cache)
 
+    def get_column_type(self, col_name: str) -> DataType:
+        for path, _type, has_subtree, _ in self.get_flat_tree():
+            if not has_subtree and DEFAULT_DELIMITER.join(path) == col_name:
+                return _type
+        raise SignalResolvingError([col_name], "is not found")
+
     def db_signals(
         self, name: Optional[str] = None, as_columns=False
     ) -> Union[list[str], list[Column]]:
@@ -490,7 +496,7 @@ def mutate(self, args_map: dict) -> "SignalSchema":
                 new_values[name] = args_map[name]
             else:
                 # adding new signal
-                new_values.update(sql_to_python({name: value}))
+                new_values[name] = sql_to_python(value)
 
         return SignalSchema(new_values)
 
@@ -534,12 +540,12 @@ def _build_tree(
             for name, val in values.items()
         }
 
-    def get_flat_tree(self) -> Iterator[tuple[list[str], type, bool, int]]:
+    def get_flat_tree(self) -> Iterator[tuple[list[str], DataType, bool, int]]:
         yield from self._get_flat_tree(self.tree, [], 0)
 
     def _get_flat_tree(
         self, tree: dict, prefix: list[str], depth: int
-    ) -> Iterator[tuple[list[str], type, bool, int]]:
+    ) -> Iterator[tuple[list[str], DataType, bool, int]]:
         for name, (type_, substree) in tree.items():
             suffix = name.split(".")
             new_prefix = prefix + suffix

src/datachain/lib/utils.py

@@ -23,3 +23,8 @@ def __init__(self, message):
 class DataChainParamsError(DataChainError):
     def __init__(self, message):
         super().__init__(message)
+
+
+class DataChainColumnError(DataChainParamsError):
+    def __init__(self, col_name, msg):
+        super().__init__(f"Error for column {col_name}: {msg}")