# Copyright (C) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License. See LICENSE in project root for information.
import sys
if sys.version >= '3':
basestring = str
from pyspark import SparkContext, SQLContext
from pyspark.sql import DataFrame
from pyspark.ml.param.shared import *
from pyspark import keyword_only
from pyspark.ml.util import JavaMLReadable, JavaMLWritable
from synapse.ml.core.platform import running_on_synapse_internal
from synapse.ml.core.serialize.java_params_patch import *
from pyspark.ml.wrapper import JavaTransformer, JavaEstimator, JavaModel
from pyspark.ml.evaluation import JavaEvaluator
from pyspark.ml.common import inherit_doc
from synapse.ml.core.schema.Utils import *
from pyspark.ml.param import TypeConverters
from synapse.ml.core.schema.TypeConversionUtils import generateTypeConverter, complexTypeConverter
[docs]@inherit_doc
class DetectLastAnomaly(ComplexParamsMixin, JavaMLReadable, JavaMLWritable, JavaTransformer):
"""
Args:
AADToken (object): AAD Token used for authentication
concurrency (int): max number of concurrent calls
concurrentTimeout (float): max number seconds to wait on futures if concurrency >= 1
customInterval (object): Custom Interval is used to set non-standard time interval, for example, if the series is 5 minutes, request can be set as granularity=minutely, customInterval=5.
errorCol (str): column to hold http errors
granularity (object): Can only be one of yearly, monthly, weekly, daily, hourly or minutely. Granularity is used for verify whether input series is valid.
handler (object): Which strategy to use when handling requests
imputeFixedValue (object): Optional argument, fixed value to use when imputeMode is set to "fixed"
imputeMode (object): Optional argument, impute mode of a time series. Possible values: auto, previous, linear, fixed, zero, notFill
maxAnomalyRatio (object): Optional argument, advanced model parameter, max anomaly ratio in a time series.
outputCol (str): The name of the output column
period (object): Optional argument, periodic value of a time series. If the value is null or does not present, the API will determine the period automatically.
sensitivity (object): Optional argument, advanced model parameter, between 0-99, the lower the value is, the larger the margin value will be which means less anomalies will be accepted
series (object): Time series data points. Points should be sorted by timestamp in ascending order to match the anomaly detection result. If the data is not sorted correctly or there is duplicated timestamp, the API will not work. In such case, an error message will be returned.
subscriptionKey (object): the API key to use
timeout (float): number of seconds to wait before closing the connection
url (str): Url of the service
"""
AADToken = Param(Params._dummy(), "AADToken", "ServiceParam: AAD Token used for authentication")
concurrency = Param(Params._dummy(), "concurrency", "max number of concurrent calls", typeConverter=TypeConverters.toInt)
concurrentTimeout = Param(Params._dummy(), "concurrentTimeout", "max number seconds to wait on futures if concurrency >= 1", typeConverter=TypeConverters.toFloat)
customInterval = Param(Params._dummy(), "customInterval", "ServiceParam: Custom Interval is used to set non-standard time interval, for example, if the series is 5 minutes, request can be set as granularity=minutely, customInterval=5. ")
errorCol = Param(Params._dummy(), "errorCol", "column to hold http errors", typeConverter=TypeConverters.toString)
granularity = Param(Params._dummy(), "granularity", "ServiceParam: Can only be one of yearly, monthly, weekly, daily, hourly or minutely. Granularity is used for verify whether input series is valid. ")
handler = Param(Params._dummy(), "handler", "Which strategy to use when handling requests")
imputeFixedValue = Param(Params._dummy(), "imputeFixedValue", "ServiceParam: Optional argument, fixed value to use when imputeMode is set to \"fixed\" ")
imputeMode = Param(Params._dummy(), "imputeMode", "ServiceParam: Optional argument, impute mode of a time series. Possible values: auto, previous, linear, fixed, zero, notFill ")
maxAnomalyRatio = Param(Params._dummy(), "maxAnomalyRatio", "ServiceParam: Optional argument, advanced model parameter, max anomaly ratio in a time series. ")
outputCol = Param(Params._dummy(), "outputCol", "The name of the output column", typeConverter=TypeConverters.toString)
period = Param(Params._dummy(), "period", "ServiceParam: Optional argument, periodic value of a time series. If the value is null or does not present, the API will determine the period automatically. ")
sensitivity = Param(Params._dummy(), "sensitivity", "ServiceParam: Optional argument, advanced model parameter, between 0-99, the lower the value is, the larger the margin value will be which means less anomalies will be accepted ")
series = Param(Params._dummy(), "series", "ServiceParam: Time series data points. Points should be sorted by timestamp in ascending order to match the anomaly detection result. If the data is not sorted correctly or there is duplicated timestamp, the API will not work. In such case, an error message will be returned. ")
subscriptionKey = Param(Params._dummy(), "subscriptionKey", "ServiceParam: the API key to use")
timeout = Param(Params._dummy(), "timeout", "number of seconds to wait before closing the connection", typeConverter=TypeConverters.toFloat)
url = Param(Params._dummy(), "url", "Url of the service", typeConverter=TypeConverters.toString)
@keyword_only
def __init__(
self,
java_obj=None,
AADToken=None,
AADTokenCol=None,
concurrency=1,
concurrentTimeout=None,
customInterval=None,
customIntervalCol=None,
errorCol="DetectLastAnomaly_b9f74647a778_error",
granularity=None,
granularityCol=None,
handler=None,
imputeFixedValue=None,
imputeFixedValueCol=None,
imputeMode=None,
imputeModeCol=None,
maxAnomalyRatio=None,
maxAnomalyRatioCol=None,
outputCol="DetectLastAnomaly_b9f74647a778_output",
period=None,
periodCol=None,
sensitivity=None,
sensitivityCol=None,
series=None,
seriesCol=None,
subscriptionKey=None,
subscriptionKeyCol=None,
timeout=60.0,
url=None
):
super(DetectLastAnomaly, self).__init__()
if java_obj is None:
self._java_obj = self._new_java_obj("com.microsoft.azure.synapse.ml.cognitive.anomaly.DetectLastAnomaly", self.uid)
else:
self._java_obj = java_obj
self._setDefault(concurrency=1)
self._setDefault(errorCol="DetectLastAnomaly_b9f74647a778_error")
self._setDefault(outputCol="DetectLastAnomaly_b9f74647a778_output")
self._setDefault(timeout=60.0)
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
if java_obj is None:
for k,v in kwargs.items():
if v is not None:
getattr(self, "set" + k[0].upper() + k[1:])(v)
[docs] @keyword_only
def setParams(
self,
AADToken=None,
AADTokenCol=None,
concurrency=1,
concurrentTimeout=None,
customInterval=None,
customIntervalCol=None,
errorCol="DetectLastAnomaly_b9f74647a778_error",
granularity=None,
granularityCol=None,
handler=None,
imputeFixedValue=None,
imputeFixedValueCol=None,
imputeMode=None,
imputeModeCol=None,
maxAnomalyRatio=None,
maxAnomalyRatioCol=None,
outputCol="DetectLastAnomaly_b9f74647a778_output",
period=None,
periodCol=None,
sensitivity=None,
sensitivityCol=None,
series=None,
seriesCol=None,
subscriptionKey=None,
subscriptionKeyCol=None,
timeout=60.0,
url=None
):
"""
Set the (keyword only) parameters
"""
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
return self._set(**kwargs)
[docs] @classmethod
def read(cls):
""" Returns an MLReader instance for this class. """
return JavaMMLReader(cls)
[docs] @staticmethod
def getJavaPackage():
""" Returns package name String. """
return "com.microsoft.azure.synapse.ml.cognitive.anomaly.DetectLastAnomaly"
@staticmethod
def _from_java(java_stage):
module_name=DetectLastAnomaly.__module__
module_name=module_name.rsplit(".", 1)[0] + ".DetectLastAnomaly"
return from_java(java_stage, module_name)
[docs] def setAADToken(self, value):
"""
Args:
AADToken: AAD Token used for authentication
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setAADToken(value)
return self
[docs] def setAADTokenCol(self, value):
"""
Args:
AADToken: AAD Token used for authentication
"""
self._java_obj = self._java_obj.setAADTokenCol(value)
return self
[docs] def setConcurrency(self, value):
"""
Args:
concurrency: max number of concurrent calls
"""
self._set(concurrency=value)
return self
[docs] def setConcurrentTimeout(self, value):
"""
Args:
concurrentTimeout: max number seconds to wait on futures if concurrency >= 1
"""
self._set(concurrentTimeout=value)
return self
[docs] def setCustomInterval(self, value):
"""
Args:
customInterval: Custom Interval is used to set non-standard time interval, for example, if the series is 5 minutes, request can be set as granularity=minutely, customInterval=5.
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setCustomInterval(value)
return self
[docs] def setCustomIntervalCol(self, value):
"""
Args:
customInterval: Custom Interval is used to set non-standard time interval, for example, if the series is 5 minutes, request can be set as granularity=minutely, customInterval=5.
"""
self._java_obj = self._java_obj.setCustomIntervalCol(value)
return self
[docs] def setErrorCol(self, value):
"""
Args:
errorCol: column to hold http errors
"""
self._set(errorCol=value)
return self
[docs] def setGranularity(self, value):
"""
Args:
granularity: Can only be one of yearly, monthly, weekly, daily, hourly or minutely. Granularity is used for verify whether input series is valid.
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setGranularity(value)
return self
[docs] def setGranularityCol(self, value):
"""
Args:
granularity: Can only be one of yearly, monthly, weekly, daily, hourly or minutely. Granularity is used for verify whether input series is valid.
"""
self._java_obj = self._java_obj.setGranularityCol(value)
return self
[docs] def setHandler(self, value):
"""
Args:
handler: Which strategy to use when handling requests
"""
self._set(handler=value)
return self
[docs] def setImputeFixedValue(self, value):
"""
Args:
imputeFixedValue: Optional argument, fixed value to use when imputeMode is set to "fixed"
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setImputeFixedValue(value)
return self
[docs] def setImputeFixedValueCol(self, value):
"""
Args:
imputeFixedValue: Optional argument, fixed value to use when imputeMode is set to "fixed"
"""
self._java_obj = self._java_obj.setImputeFixedValueCol(value)
return self
[docs] def setImputeMode(self, value):
"""
Args:
imputeMode: Optional argument, impute mode of a time series. Possible values: auto, previous, linear, fixed, zero, notFill
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setImputeMode(value)
return self
[docs] def setImputeModeCol(self, value):
"""
Args:
imputeMode: Optional argument, impute mode of a time series. Possible values: auto, previous, linear, fixed, zero, notFill
"""
self._java_obj = self._java_obj.setImputeModeCol(value)
return self
[docs] def setMaxAnomalyRatio(self, value):
"""
Args:
maxAnomalyRatio: Optional argument, advanced model parameter, max anomaly ratio in a time series.
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setMaxAnomalyRatio(value)
return self
[docs] def setMaxAnomalyRatioCol(self, value):
"""
Args:
maxAnomalyRatio: Optional argument, advanced model parameter, max anomaly ratio in a time series.
"""
self._java_obj = self._java_obj.setMaxAnomalyRatioCol(value)
return self
[docs] def setOutputCol(self, value):
"""
Args:
outputCol: The name of the output column
"""
self._set(outputCol=value)
return self
[docs] def setPeriod(self, value):
"""
Args:
period: Optional argument, periodic value of a time series. If the value is null or does not present, the API will determine the period automatically.
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setPeriod(value)
return self
[docs] def setPeriodCol(self, value):
"""
Args:
period: Optional argument, periodic value of a time series. If the value is null or does not present, the API will determine the period automatically.
"""
self._java_obj = self._java_obj.setPeriodCol(value)
return self
[docs] def setSensitivity(self, value):
"""
Args:
sensitivity: Optional argument, advanced model parameter, between 0-99, the lower the value is, the larger the margin value will be which means less anomalies will be accepted
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setSensitivity(value)
return self
[docs] def setSensitivityCol(self, value):
"""
Args:
sensitivity: Optional argument, advanced model parameter, between 0-99, the lower the value is, the larger the margin value will be which means less anomalies will be accepted
"""
self._java_obj = self._java_obj.setSensitivityCol(value)
return self
[docs] def setSeries(self, value):
"""
Args:
series: Time series data points. Points should be sorted by timestamp in ascending order to match the anomaly detection result. If the data is not sorted correctly or there is duplicated timestamp, the API will not work. In such case, an error message will be returned.
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setSeries(value)
return self
[docs] def setSeriesCol(self, value):
"""
Args:
series: Time series data points. Points should be sorted by timestamp in ascending order to match the anomaly detection result. If the data is not sorted correctly or there is duplicated timestamp, the API will not work. In such case, an error message will be returned.
"""
self._java_obj = self._java_obj.setSeriesCol(value)
return self
[docs] def setSubscriptionKey(self, value):
"""
Args:
subscriptionKey: the API key to use
"""
if isinstance(value, list):
value = SparkContext._active_spark_context._jvm.com.microsoft.azure.synapse.ml.param.ServiceParam.toSeq(value)
self._java_obj = self._java_obj.setSubscriptionKey(value)
return self
[docs] def setSubscriptionKeyCol(self, value):
"""
Args:
subscriptionKey: the API key to use
"""
self._java_obj = self._java_obj.setSubscriptionKeyCol(value)
return self
[docs] def setTimeout(self, value):
"""
Args:
timeout: number of seconds to wait before closing the connection
"""
self._set(timeout=value)
return self
[docs] def setUrl(self, value):
"""
Args:
url: Url of the service
"""
self._set(url=value)
return self
[docs] def getAADToken(self):
"""
Returns:
AADToken: AAD Token used for authentication
"""
return self._java_obj.getAADToken()
[docs] def getConcurrency(self):
"""
Returns:
concurrency: max number of concurrent calls
"""
return self.getOrDefault(self.concurrency)
[docs] def getConcurrentTimeout(self):
"""
Returns:
concurrentTimeout: max number seconds to wait on futures if concurrency >= 1
"""
return self.getOrDefault(self.concurrentTimeout)
[docs] def getCustomInterval(self):
"""
Returns:
customInterval: Custom Interval is used to set non-standard time interval, for example, if the series is 5 minutes, request can be set as granularity=minutely, customInterval=5.
"""
return self._java_obj.getCustomInterval()
[docs] def getErrorCol(self):
"""
Returns:
errorCol: column to hold http errors
"""
return self.getOrDefault(self.errorCol)
[docs] def getGranularity(self):
"""
Returns:
granularity: Can only be one of yearly, monthly, weekly, daily, hourly or minutely. Granularity is used for verify whether input series is valid.
"""
return self._java_obj.getGranularity()
[docs] def getHandler(self):
"""
Returns:
handler: Which strategy to use when handling requests
"""
return self.getOrDefault(self.handler)
[docs] def getImputeFixedValue(self):
"""
Returns:
imputeFixedValue: Optional argument, fixed value to use when imputeMode is set to "fixed"
"""
return self._java_obj.getImputeFixedValue()
[docs] def getImputeMode(self):
"""
Returns:
imputeMode: Optional argument, impute mode of a time series. Possible values: auto, previous, linear, fixed, zero, notFill
"""
return self._java_obj.getImputeMode()
[docs] def getMaxAnomalyRatio(self):
"""
Returns:
maxAnomalyRatio: Optional argument, advanced model parameter, max anomaly ratio in a time series.
"""
return self._java_obj.getMaxAnomalyRatio()
[docs] def getOutputCol(self):
"""
Returns:
outputCol: The name of the output column
"""
return self.getOrDefault(self.outputCol)
[docs] def getPeriod(self):
"""
Returns:
period: Optional argument, periodic value of a time series. If the value is null or does not present, the API will determine the period automatically.
"""
return self._java_obj.getPeriod()
[docs] def getSensitivity(self):
"""
Returns:
sensitivity: Optional argument, advanced model parameter, between 0-99, the lower the value is, the larger the margin value will be which means less anomalies will be accepted
"""
return self._java_obj.getSensitivity()
[docs] def getSeries(self):
"""
Returns:
series: Time series data points. Points should be sorted by timestamp in ascending order to match the anomaly detection result. If the data is not sorted correctly or there is duplicated timestamp, the API will not work. In such case, an error message will be returned.
"""
return self._java_obj.getSeries()
[docs] def getSubscriptionKey(self):
"""
Returns:
subscriptionKey: the API key to use
"""
return self._java_obj.getSubscriptionKey()
[docs] def getTimeout(self):
"""
Returns:
timeout: number of seconds to wait before closing the connection
"""
return self.getOrDefault(self.timeout)
[docs] def getUrl(self):
"""
Returns:
url: Url of the service
"""
return self.getOrDefault(self.url)
[docs] def setCustomServiceName(self, value):
self._java_obj = self._java_obj.setCustomServiceName(value)
return self
[docs] def setEndpoint(self, value):
self._java_obj = self._java_obj.setEndpoint(value)
return self
[docs] def setDefaultInternalEndpoint(self, value):
self._java_obj = self._java_obj.setDefaultInternalEndpoint(value)
return self
def _transform(self, dataset: DataFrame) -> DataFrame:
if running_on_synapse_internal():
try:
from synapse.ml.mlflow import get_mlflow_env_config
mlflow_env_configs = get_mlflow_env_config()
self._java_obj.setDefaultAADToken(mlflow_env_configs.driver_aad_token)
self.setDefaultInternalEndpoint(mlflow_env_configs.workload_endpoint)
except ModuleNotFoundError as e:
pass
return super()._transform(dataset)
[docs] def setLocation(self, value):
self._java_obj = self._java_obj.setLocation(value)
return self
[docs] def setLinkedService(self, value):
self._java_obj = self._java_obj.setLinkedService(value)
return self