def test_model_log_with_databricks_runtime(): dbr = "8.3.x-snapshot-gpu-ml-scala2.12" with TempDir(chdr=True) as tmp, mock.patch( "mlflow.models.model.get_databricks_runtime", return_value=dbr): sig = ModelSignature( inputs=Schema([ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")]), ) input_example = {"x": 1, "y": 2} local_path, r = _log_model_with_signature_and_example( tmp, sig, input_example) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) assert loaded_model.run_id == r.info.run_id assert loaded_model.artifact_path == "some/path" assert loaded_model.flavors == { "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, } assert loaded_model.signature == sig path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path) assert x.to_dict(orient="records")[0] == input_example assert loaded_model.databricks_runtime == dbr
def _dataframe_from_json(path_or_str, schema: Schema = None, pandas_orient: str = "split") -> pd.DataFrame: """ Parse json into pandas.DataFrame. User can pass schema to ensure correct type parsing and to make any necessary conversions (e.g. string -> binary for binary columns). :param path_or_str: Path to a json file or a json string. :param schema: Mlflow schema used when parsing the data. :param pandas_orient: pandas data frame convention used to store the data. :return: pandas.DataFrame. """ if schema is not None: dtypes = dict(zip(schema.column_names(), schema.column_types())) df = pd.read_json(path_or_str, orient=pandas_orient, dtype=dtypes)[schema.column_names()] binary_cols = [ i for i, x in enumerate(schema.column_types()) if x == DataType.binary ] for i in binary_cols: col = df.columns[i] df[col] = np.array(df[col].map(_base64decode), dtype=np.bytes_) return df else: return pd.read_json(path_or_str, orient=pandas_orient, dtype=False)
def test_model_log(): with TempDir(chdr=True) as tmp: sig = ModelSignature( inputs=Schema([ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")]), ) input_example = {"x": 1, "y": 2} local_path, r = _log_model_with_signature_and_example( tmp, sig, input_example) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) assert loaded_model.run_id == r.info.run_id assert loaded_model.artifact_path == "some/path" assert loaded_model.flavors == { "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, } assert loaded_model.signature == sig path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path) assert x.to_dict(orient="records")[0] == input_example assert not hasattr(loaded_model, "databricks_runtime")
def on_train_end(self, args, state, control, **kwargs): input_schema = Schema([ColSpec(name="text", type="string")]) output_schema = Schema([TensorSpec(np.dtype(np.float), (-1, -1))]) signature = ModelSignature(inputs=input_schema, outputs=output_schema) pyfunc.log_model( # artifact path is _relative_ to run root in mlflow artifact_path="bert_classifier_model", # Dir with the module files for dependencies code_path=[ os.path.join(os.path.dirname(os.path.abspath(__file__)), "models.py"), os.path.join(os.path.dirname(os.path.abspath(__file__)), "utils.py") ], python_model=MLFlowBertClassificationModel(), artifacts={ "model": state.best_model_checkpoint, }, conda_env={ 'name': 'classifier-env', 'channels': ['defaults', 'pytorch', 'pypi'], 'dependencies': [ 'python=3.8.8', 'pip', 'pytorch=1.8.0', { 'pip': [ 'transformers==4.4.2', 'mlflow==1.15.0', 'numpy==1.20.1' ] } ] }, signature=signature, await_registration_for=5, registered_model_name=self.registered_name)
def test_model_load_input_example_failures(): with TempDir(chdr=True) as tmp: input_example = np.array([[3, 4, 5]], dtype=np.int32) sig = ModelSignature( inputs=Schema([ TensorSpec(type=input_example.dtype, shape=input_example.shape) ]), outputs=Schema([ColSpec(name=None, type="double")]), ) local_path, _ = _log_model_with_signature_and_example( tmp, sig, input_example) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) loaded_example = loaded_model.load_input_example(local_path) assert loaded_example is not None with pytest.raises(FileNotFoundError, match="No such file or directory"): loaded_model.load_input_example( os.path.join(local_path, "folder_which_does_not_exist")) path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) os.remove(path) with pytest.raises(FileNotFoundError, match="No such file or directory"): loaded_model.load_input_example(local_path)
def test_spark_schema_inference(pandas_df_with_all_types): import pyspark from pyspark.sql.types import _parse_datatype_string, StructField, StructType pandas_df_with_all_types = pandas_df_with_all_types.drop( columns=["boolean_ext", "integer_ext", "string_ext"]) schema = _infer_schema(pandas_df_with_all_types) assert schema == Schema( [ColSpec(x, x) for x in pandas_df_with_all_types.columns]) spark_session = pyspark.sql.SparkSession( pyspark.SparkContext.getOrCreate()) struct_fields = [] for t in schema.input_types(): # pyspark _parse_datatype_string() expects "timestamp" instead of "datetime" if t == DataType.datetime: struct_fields.append( StructField("datetime", _parse_datatype_string("timestamp"), True)) else: struct_fields.append( StructField(t.name, _parse_datatype_string(t.name), True)) spark_schema = StructType(struct_fields) sparkdf = spark_session.createDataFrame(pandas_df_with_all_types, schema=spark_schema) schema = _infer_schema(sparkdf) assert schema == Schema( [ColSpec(x, x) for x in pandas_df_with_all_types.columns])
def test_model_log(): with TempDir(chdr=True) as tmp: experiment_id = mlflow.create_experiment("test") sig = ModelSignature( inputs=Schema([ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")])) input_example = {"x": 1, "y": 2} with mlflow.start_run(experiment_id=experiment_id) as r: Model.log("some/path", TestFlavor, signature=sig, input_example=input_example) local_path = _download_artifact_from_uri("runs:/{}/some/path".format( r.info.run_id), output_path=tmp.path("")) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) assert loaded_model.run_id == r.info.run_id assert loaded_model.artifact_path == "some/path" assert loaded_model.flavors == { "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, } assert loaded_model.signature == sig path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path) assert x.to_dict(orient="records")[0] == input_example
def _dataframe_from_json(path_or_str, schema: Schema = None, pandas_orient: str = "split", precise_float=False) -> pd.DataFrame: """ Parse json into pandas.DataFrame. User can pass schema to ensure correct type parsing and to make any necessary conversions (e.g. string -> binary for binary columns). :param path_or_str: Path to a json file or a json string. :param schema: Mlflow schema used when parsing the data. :param pandas_orient: pandas data frame convention used to store the data. :return: pandas.DataFrame. """ if schema is not None: dtypes = dict(zip(schema.column_names(), schema.pandas_types())) df = pd.read_json(path_or_str, orient=pandas_orient, dtype=dtypes, precise_float=precise_float) actual_cols = set(df.columns) for type_, name in zip(schema.column_types(), schema.column_names()): if type_ == DataType.binary and name in actual_cols: df[name] = df[name].map( lambda x: base64.decodebytes(bytes(x, "utf8"))) return df else: return pd.read_json(path_or_str, orient=pandas_orient, dtype=False, precise_float=precise_float)
def test_model_log_with_input_example_succeeds(): with TempDir(chdr=True) as tmp: sig = ModelSignature( inputs=Schema([ ColSpec("integer", "a"), ColSpec("string", "b"), ColSpec("boolean", "c"), ColSpec("string", "d"), ColSpec("datetime", "e"), ]), outputs=Schema([ColSpec(name=None, type="double")]), ) input_example = pd.DataFrame( { "a": np.int32(1), "b": "test string", "c": True, "d": date.today(), "e": np.datetime64("2020-01-01T00:00:00"), }, index=[0], ) local_path, _ = _log_model_with_signature_and_example( tmp, sig, input_example) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path, schema=sig.inputs) # date column will get deserialized into string input_example["d"] = input_example["d"].apply(lambda x: x.isoformat()) assert x.equals(input_example)
def train( proj_name: str, Model: str, dataset_cls: str, net_fn: str, net_args: Dict, dataset_args: Dict, ): """ Train Function """ dataset_module = importlib.import_module( f"manythings.data.dta_{dataset_cls}") dataset_cls_ = getattr(dataset_module, dataset_cls) network_module = importlib.import_module(f"manythings.networks.{net_fn}") network_fn_ = getattr(network_module, net_fn) model_module = importlib.import_module(f"manythings.models.{Model}") model_cls_ = getattr(model_module, Model) config = { "model": Model, "dataset_cls": dataset_cls, "net_fn": net_fn, "net_args": net_args, "dataset_args": dataset_args } input_schema = Schema([ TensorSpec(np.dtype(np.uint8), (-1, 71), "encoder_input"), TensorSpec(np.dtype(np.uint8), (-1, 93), "decoder_input") ]) output_schema = Schema([TensorSpec(np.dtype(np.float32), (-1, 93))]) signature = ModelSignature(inputs=input_schema, outputs=output_schema) data = dataset_cls_() data.load_or_generate() data.preprocess() with wandb.init(project=proj_name, config=config): """""" config = wandb.config model = model_cls_(dataset_cls_, network_fn_, net_args, dataset_args) callbacks = [ WandbCallback( # training_data=( # [data.encoder_input_data, data.decoder_input_data], # data.decoder_target_data # ), # log_weights=True, # log_gradients=True ) ] model.fit(callbacks=callbacks) mlflow.keras.save_model(model.network, "saved_models/seq2seq", signature=signature)
def test_model_save_load(): m = Model(artifact_path="some/path", run_id="123", flavors={ "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, }, signature=ModelSignature( inputs=Schema( [ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")])), saved_input_example_info={ "x": 1, "y": 2 }) assert m.get_input_schema() == m.signature.inputs assert m.get_output_schema() == m.signature.outputs x = Model(artifact_path="some/other/path", run_id="1234") assert x.get_input_schema() is None assert x.get_output_schema() is None n = Model(artifact_path="some/path", run_id="123", flavors={ "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, }, signature=ModelSignature( inputs=Schema( [ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")])), saved_input_example_info={ "x": 1, "y": 2 }) n.utc_time_created = m.utc_time_created assert m == n n.signature = None assert m != n with TempDir() as tmp: m.save(tmp.path("model")) o = Model.load(tmp.path("model")) assert m == o assert m.to_json() == o.to_json() assert m.to_yaml() == o.to_yaml()
def test_schema_creation_with_named_and_unnamed_spec(): with pytest.raises(MlflowException) as ex: Schema([ TensorSpec(np.dtype("float64"), (-1, ), "blah"), TensorSpec(np.dtype("float64"), (-1, )) ]) assert "Creating Schema with a combination of named and unnamed columns" in ex.value.message with pytest.raises(MlflowException) as ex: Schema([ColSpec("double", "blah"), ColSpec("double")]) assert "Creating Schema with a combination of named and unnamed columns" in ex.value.message
def test_model_load_input_example_no_signature(): with TempDir(chdr=True) as tmp: input_example = np.array([[3, 4, 5]], dtype=np.int32) sig = ModelSignature( inputs=Schema([TensorSpec(type=input_example.dtype, shape=input_example.shape)]), outputs=Schema([ColSpec(name=None, type="double")]), ) local_path, _ = _log_model_with_signature_and_example(tmp, sig, input_example=None) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) loaded_example = loaded_model.load_input_example(local_path) assert loaded_example is None
def test_model_load_input_example_scipy(): with TempDir(chdr=True) as tmp: input_example = csc_matrix(np.arange(0, 12, 0.5).reshape(3, 8)) sig = ModelSignature( inputs=Schema([TensorSpec(type=input_example.data.dtype, shape=input_example.shape)]), outputs=Schema([ColSpec(name=None, type="double")]), ) local_path, _ = _log_model_with_signature_and_example(tmp, sig, input_example) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) loaded_example = loaded_model.load_input_example(local_path) assert isinstance(loaded_example, csc_matrix) assert np.array_equal(input_example.data, loaded_example.data)
def test_schema_creation(): # can create schema with named col specs Schema([ColSpec("double", "a"), ColSpec("integer", "b")]) # can create schema with unnamed col specs Schema([ColSpec("double"), ColSpec("integer")]) # can create schema with multiple named tensor specs Schema([TensorSpec(np.dtype("float64"), (-1,), "a"), TensorSpec(np.dtype("uint8"), (-1,), "b")]) # can create schema with single unnamed tensor spec Schema([TensorSpec(np.dtype("float64"), (-1,))]) # combination of tensor and col spec is not allowed with pytest.raises(MlflowException) as ex: Schema([TensorSpec(np.dtype("float64"), (-1,)), ColSpec("double")]) assert "Please choose one of" in ex.value.message # combination of named and unnamed inputs is not allowed with pytest.raises(MlflowException) as ex: Schema( [TensorSpec(np.dtype("float64"), (-1,), "blah"), TensorSpec(np.dtype("float64"), (-1,))] ) assert "Creating Schema with a combination of named and unnamed inputs" in ex.value.message with pytest.raises(MlflowException) as ex: Schema([ColSpec("double", "blah"), ColSpec("double")]) assert "Creating Schema with a combination of named and unnamed inputs" in ex.value.message # multiple unnamed tensor specs is not allowed with pytest.raises(MlflowException) as ex: Schema([TensorSpec(np.dtype("double"), (-1,)), TensorSpec(np.dtype("double"), (-1,))]) assert "Creating Schema with multiple unnamed TensorSpecs is not supported" in ex.value.message
def test_model_info(): with TempDir(chdr=True) as tmp: sig = ModelSignature( inputs=Schema([ColSpec("integer", "x"), ColSpec("integer", "y")]), outputs=Schema([ColSpec(name=None, type="double")]), ) input_example = {"x": 1, "y": 2} experiment_id = mlflow.create_experiment("test") with mlflow.start_run(experiment_id=experiment_id) as run: model_info = Model.log("some/path", TestFlavor, signature=sig, input_example=input_example) local_path = _download_artifact_from_uri("runs:/{}/some/path".format( run.info.run_id), output_path=tmp.path("")) assert model_info.run_id == run.info.run_id assert model_info.artifact_path == "some/path" assert model_info.model_uri == "runs:/{}/some/path".format( run.info.run_id) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) assert model_info.utc_time_created == loaded_model.utc_time_created assert model_info.model_uuid == loaded_model.model_uuid assert model_info.flavors == { "flavor1": { "a": 1, "b": 2 }, "flavor2": { "x": 1, "y": 2 }, } path = os.path.join( local_path, model_info.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path) assert x.to_dict(orient="records")[0] == input_example assert model_info.signature_dict == sig.to_dict() assert Version(model_info.mlflow_version) == Version( loaded_model.mlflow_version)
def test_spark_schema_inference(pandas_df_with_all_types): import pyspark from pyspark.sql.types import _parse_datatype_string, StructField, StructType pandas_df_with_all_types = pandas_df_with_all_types.drop( columns=["boolean_ext", "integer_ext", "string_ext"] ) schema = _infer_schema(pandas_df_with_all_types) assert schema == Schema([ColSpec(x, x) for x in pandas_df_with_all_types.columns]) spark_session = pyspark.sql.SparkSession(pyspark.SparkContext.getOrCreate()) spark_schema = StructType( [StructField(t.name, _parse_datatype_string(t.name), True) for t in schema.column_types()] ) sparkdf = spark_session.createDataFrame(pandas_df_with_all_types, schema=spark_schema) schema = _infer_schema(sparkdf) assert schema == Schema([ColSpec(x, x) for x in pandas_df_with_all_types.columns])
def test_schema_inference_on_dataframe(pandas_df_with_all_types): basic_types = pandas_df_with_all_types.drop( columns=["boolean_ext", "integer_ext", "string_ext"]) schema = _infer_schema(basic_types) assert schema == Schema([ColSpec(x, x) for x in basic_types.columns]) ext_types = pandas_df_with_all_types[[ "boolean_ext", "integer_ext", "string_ext" ]].copy() expected_schema = Schema([ ColSpec(DataType.boolean, "boolean_ext"), ColSpec(DataType.long, "integer_ext"), ColSpec(DataType.string, "string_ext"), ]) schema = _infer_schema(ext_types) assert schema == expected_schema
def from_dict(cls, signature_dict: Dict[str, Any]): """ Deserialize from dictionary representation. :param signature_dict: Dictionary representation of model signature. Expected dictionary format: `{'inputs': <json string>, 'outputs': <json string>" }` :return: ModelSignature populated with the data form the dictionary. """ inputs = Schema.from_json(signature_dict["inputs"]) if "outputs" in signature_dict and signature_dict["outputs"] is not None: outputs = Schema.from_json(signature_dict["outputs"]) return cls(inputs, outputs) else: return cls(inputs)
def test_dtype(nparray, dtype): schema = _infer_schema(nparray) assert schema == Schema([TensorSpec(np.dtype(dtype), (-1, ))]) spec = schema.inputs[0] recreated_spec = TensorSpec.from_json_dict(**spec.to_dict()) assert spec == recreated_spec enforced_array = _enforce_tensor_spec(nparray, spec) assert isinstance(enforced_array, np.ndarray)
def test_model_signature(): signature1 = ModelSignature(inputs=Schema( [ColSpec(DataType.boolean), ColSpec(DataType.binary)]), outputs=Schema([ ColSpec(name=None, type=DataType.double), ColSpec(name=None, type=DataType.double) ])) signature2 = ModelSignature(inputs=Schema( [ColSpec(DataType.boolean), ColSpec(DataType.binary)]), outputs=Schema([ ColSpec(name=None, type=DataType.double), ColSpec(name=None, type=DataType.double) ])) assert signature1 == signature2 signature3 = ModelSignature(inputs=Schema( [ColSpec(DataType.boolean), ColSpec(DataType.binary)]), outputs=Schema([ ColSpec(name=None, type=DataType.float), ColSpec(name=None, type=DataType.double) ])) assert signature3 != signature1 as_json = json.dumps(signature1.to_dict()) signature4 = ModelSignature.from_dict(json.loads(as_json)) assert signature1 == signature4 signature5 = ModelSignature(inputs=Schema( [ColSpec(DataType.boolean), ColSpec(DataType.binary)]), outputs=None) as_json = json.dumps(signature5.to_dict()) signature6 = ModelSignature.from_dict(json.loads(as_json)) assert signature5 == signature6
def test_content_types(tensor_spec: TensorSpec, request_input: RequestInput): input_schema = Schema(inputs=[tensor_spec]) inference_request = InferenceRequest( parameters=Parameters(content_type=PandasCodec.ContentType), inputs=[request_input], ) data = decode_inference_request(inference_request) # _enforce_schema will raise if something fails _enforce_schema(data, input_schema)
def test_signature_inference_infers_datime_types_as_expected(): col_name = "datetime_col" test_datetime = np.datetime64("2021-01-01") test_series = pd.Series(pd.to_datetime([test_datetime])) test_df = test_series.to_frame(col_name) signature = infer_signature(test_series) assert signature.inputs == Schema([ColSpec(DataType.datetime)]) signature = infer_signature(test_df) assert signature.inputs == Schema([ColSpec(DataType.datetime, name=col_name)]) spark = pyspark.sql.SparkSession.builder.getOrCreate() spark_df = spark.range(1).selectExpr( "current_timestamp() as timestamp", "current_date() as date" ) signature = infer_signature(spark_df) assert signature.inputs == Schema( [ColSpec(DataType.datetime, name="timestamp"), ColSpec(DataType.datetime, name="date")] )
def test_schema_inference_on_dictionary(dict_of_ndarrays): # test dictionary schema = _infer_schema(dict_of_ndarrays) assert schema == Schema([ TensorSpec(tensor.dtype, _get_tensor_shape(tensor), name) for name, tensor in dict_of_ndarrays.items() ]) # test exception is raised if non-numpy data in dictionary with pytest.raises(TypeError): _infer_schema({"x": 1}) with pytest.raises(TypeError): _infer_schema({"x": [1]})
def test_model_log_with_input_example_succeeds(): with TempDir(chdr=True) as tmp: experiment_id = mlflow.create_experiment("test") sig = ModelSignature( inputs=Schema([ ColSpec("integer", "a"), ColSpec("string", "b"), ColSpec("boolean", "c"), ColSpec("string", "d"), ColSpec("datetime", "e"), ]), outputs=Schema([ColSpec(name=None, type="double")]), ) input_example = pd.DataFrame( { "a": np.int32(1), "b": "test string", "c": True, "d": date.today(), "e": np.datetime64("2020-01-01T00:00:00"), }, index=[0], ) with mlflow.start_run(experiment_id=experiment_id) as r: Model.log("some/path", TestFlavor, signature=sig, input_example=input_example) local_path = _download_artifact_from_uri("runs:/{}/some/path".format( r.info.run_id), output_path=tmp.path("")) loaded_model = Model.load(os.path.join(local_path, "MLmodel")) path = os.path.join( local_path, loaded_model.saved_input_example_info["artifact_path"]) x = _dataframe_from_json(path, schema=sig.inputs) # date column will get deserialized into string input_example["d"] = input_example["d"].apply(lambda x: x.isoformat()) assert x.equals(input_example)
def test_schema_inference_on_numpy_array(pandas_df_with_all_types): for col in pandas_df_with_all_types: data = pandas_df_with_all_types[col].to_numpy() schema = _infer_schema(data) assert schema == Schema([TensorSpec(type=data.dtype, shape=(-1, ))]) # test boolean schema = _infer_schema(np.array([True, False, True], dtype=np.bool_)) assert schema == Schema([TensorSpec(np.dtype(np.bool_), (-1, ))]) # test bytes schema = _infer_schema(np.array([bytes([1])], dtype=np.bytes_)) assert schema == Schema([TensorSpec(np.dtype("S1"), (-1, ))]) # test (u)ints for t in [ np.uint8, np.int8, np.uint16, np.int16, np.uint32, np.int32, np.uint64, np.int64 ]: schema = _infer_schema(np.array([1, 2, 3], dtype=t)) assert schema == Schema([TensorSpec(np.dtype(t), (-1, ))]) # test floats for t in [np.float16, np.float32, np.float64]: schema = _infer_schema(np.array([1.1, 2.2, 3.3], dtype=t)) assert schema == Schema([TensorSpec(np.dtype(t), (-1, ))]) if hasattr(np, "float128"): schema = _infer_schema(np.array([1.1, 2.2, 3.3], dtype=np.float128)) assert schema == Schema([TensorSpec(np.dtype(np.float128), (-1, ))])
def test_spark_type_mapping(pandas_df_with_all_types): import pyspark from pyspark.sql.types import ( BooleanType, IntegerType, LongType, FloatType, DoubleType, StringType, BinaryType, TimestampType, ) from pyspark.sql.types import StructField, StructType assert isinstance(DataType.boolean.to_spark(), BooleanType) assert isinstance(DataType.integer.to_spark(), IntegerType) assert isinstance(DataType.long.to_spark(), LongType) assert isinstance(DataType.float.to_spark(), FloatType) assert isinstance(DataType.double.to_spark(), DoubleType) assert isinstance(DataType.string.to_spark(), StringType) assert isinstance(DataType.binary.to_spark(), BinaryType) assert isinstance(DataType.datetime.to_spark(), TimestampType) pandas_df_with_all_types = pandas_df_with_all_types.drop( columns=["boolean_ext", "integer_ext", "string_ext"]) schema = _infer_schema(pandas_df_with_all_types) expected_spark_schema = StructType([ StructField(t.name, t.to_spark(), True) for t in schema.input_types() ]) actual_spark_schema = schema.as_spark_schema() assert expected_spark_schema.jsonValue() == actual_spark_schema.jsonValue() spark_session = pyspark.sql.SparkSession( pyspark.SparkContext.getOrCreate()) sparkdf = spark_session.createDataFrame(pandas_df_with_all_types, schema=actual_spark_schema) schema2 = _infer_schema(sparkdf) assert schema == schema2 # test unnamed columns schema = Schema([ColSpec(col.type) for col in schema.inputs]) expected_spark_schema = StructType([ StructField(str(i), t.to_spark(), True) for i, t in enumerate(schema.input_types()) ]) actual_spark_schema = schema.as_spark_schema() assert expected_spark_schema.jsonValue() == actual_spark_schema.jsonValue() # test single unnamed column is mapped to just a single spark type schema = Schema([ColSpec(DataType.integer)]) spark_type = schema.as_spark_schema() assert isinstance(spark_type, IntegerType)
def _infer_schema(data): res = [] for _, col in enumerate(data): t = col.type.replace("tensor(", "").replace(")", "") if t in ["bool"]: dt = DataType.boolean elif t in ["int8", "uint8", "int16", "uint16", "int32"]: dt = DateType.integer elif t in ["uint32", "int64"]: dt = DataType.long elif t in ["float16", "bfloat16", "float"]: dt = DataType.float elif t in ["double"]: dt = DataType.double elif t in ["string"]: dt = DataType.string else: raise Exception("Unsupported type: " + t) res.append(ColSpec(type=dt, name=col.name)) return Schema(res)
def test_model_signature_with_colspec_and_tensorspec(): signature1 = ModelSignature(inputs=Schema([ColSpec(DataType.double)])) signature2 = ModelSignature(inputs=Schema([TensorSpec(np.dtype("float"), (-1, 28, 28))])) assert signature1 != signature2 assert signature2 != signature1 signature3 = ModelSignature( inputs=Schema([ColSpec(DataType.double)]), outputs=Schema([TensorSpec(np.dtype("float"), (-1, 28, 28))]), ) signature4 = ModelSignature( inputs=Schema([ColSpec(DataType.double)]), outputs=Schema([ColSpec(DataType.double)]), ) assert signature3 != signature4 assert signature4 != signature3
def _infer_schema(data: Any) -> Schema: """ Infer an MLflow schema from a dataset. This method captures the column names and data types from the user data. The signature represents model input and output as data frames with (optionally) named columns and data type specified as one of types defined in :py:class:`DataType`. This method will raise an exception if the user data contains incompatible types or is not passed in one of the supported formats (containers). The input should be one of these: - pandas.DataFrame or pandas.Series - dictionary of { name -> numpy.ndarray} - numpy.ndarray - pyspark.sql.DataFrame The element types should be mappable to one of :py:class:`mlflow.models.signature.DataType`. NOTE: Multidimensional (>2d) arrays (aka tensors) are not supported at this time. :param data: Dataset to infer from. :return: Schema """ if isinstance(data, dict): res = [] for col in data.keys(): ary = data[col] if not isinstance(ary, np.ndarray): raise TypeError( "Data in the dictionary must be of type numpy.ndarray") dims = len(ary.shape) if dims == 1: res.append(ColSpec(type=_infer_numpy_array(ary), name=col)) else: raise TensorsNotSupportedException( "Data in the dictionary must be 1-dimensional, " "got shape {}".format(ary.shape)) return Schema(res) elif isinstance(data, pd.Series): return Schema([ColSpec(type=_infer_numpy_array(data.values))]) elif isinstance(data, pd.DataFrame): return Schema([ ColSpec(type=_infer_numpy_array(data[col].values), name=col) for col in data.columns ]) elif isinstance(data, np.ndarray): if len(data.shape) > 2: raise TensorsNotSupportedException( "Attempting to infer schema from numpy array with " "shape {}".format(data.shape)) if data.dtype == np.object: data = pd.DataFrame(data).infer_objects() return Schema([ ColSpec(type=_infer_numpy_array(data[col].values)) for col in data.columns ]) if len(data.shape) == 1: return Schema([ColSpec(type=_infer_numpy_dtype(data.dtype))]) elif len(data.shape) == 2: return Schema([ColSpec(type=_infer_numpy_dtype(data.dtype))] * data.shape[1]) elif _is_spark_df(data): return Schema([ ColSpec(type=_infer_spark_type(field.dataType), name=field.name) for field in data.schema.fields ]) raise TypeError( "Expected one of (pandas.DataFrame, numpy array, " "dictionary of (name -> numpy.ndarray), pyspark.sql.DataFrame) " "but got '{}'".format(type(data)))