Pyspark ML - 如何保存管道和 RandomForestClassificationModel

我无法保存使用 python/spark 的 ml 包生成的随机森林模型。

>>> rf = RandomForestClassifier(labelCol="label", featuresCol="features")
>>> pipeline = Pipeline(stages=early_stages + [rf])
>>> model = pipeline.fit(trainingData)
>>> model.save("fittedpipeline")

回溯（最近一次调用）：文件“”，第 1 行，位于 AttributeError：“PipelineModel”对象没有属性 '节省'

>>> rfModel = model.stages[8]
>>> print(rfModel)

RandomForestClassificationModel (uid=rfc_46c07f6d7ac8) 有 20 棵树

>> rfModel.save("rfmodel")

回溯（最近一次调用）：文件“”，第 1 行，位于 属性错误：“RandomForestClassificationModel”对象有 没有属性“保存”**

还尝试通过传递“sc”作为保存方法的第一个参数。

您的代码的主要问题是您使用的是 2.0.0 之前的 Apache Spark 版本。因此，save尚不可用Pipeline API.

这是根据官方文档合成的完整示例。让我们首先创建我们的管道：

from pyspark.ml import Pipeline
from pyspark.ml.classification import RandomForestClassifier
from pyspark.ml.feature import IndexToString, StringIndexer, VectorIndexer

# Load and parse the data file, converting it to a DataFrame.
data = spark.read.format("libsvm").load("data/mllib/sample_libsvm_data.txt")

# Index labels, adding metadata to the label column.
# Fit on whole dataset to include all labels in index.
label_indexer = StringIndexer(inputCol="label", outputCol="indexedLabel")
labels = label_indexer.fit(data).labels

# Automatically identify categorical features, and index them.
# Set maxCategories so features with > 4 distinct values are treated as continuous.
feature_indexer = VectorIndexer(inputCol="features", outputCol="indexedFeatures", maxCategories=4)

early_stages = [label_indexer, feature_indexer]

# Split the data into training and test sets (30% held out for testing)
(train, test) = data.randomSplit([0.7, 0.3])

# Train a RandomForest model.
rf = RandomForestClassifier(labelCol="indexedLabel", featuresCol="indexedFeatures", numTrees=10)

# Convert indexed labels back to original labels.
label_converter = IndexToString(inputCol="prediction", outputCol="predictedLabel", labels=labels)

# Chain indexers and forest in a Pipeline
pipeline = Pipeline(stages=early_stages + [rf, label_converter])

# Train model. This also runs the indexers.
model = pipeline.fit(train)

您现在可以保存管道：

>>> model.save("/tmp/rf")
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.

您还可以保存 RF 模型：

>>> rf_model = model.stages[2]
>>> print(rf_model)
RandomForestClassificationModel (uid=rfc_b368678f4122) with 10 trees
>>> rf_model.save("/tmp/rf_2")