PYSPARK: how to get weights from CrossValidatorModel? - apache-spark

Question

I have trained a logistic regression model using cross validation using the following code from https://spark.apache.org/docs/2.1.0/ml-tuning.html
now I want to get the weights and intercept, but I get this error:
AttributeError: 'CrossValidatorModel' object has no attribute 'weights'
how can I get these attributes?
*the same problem with (trainingSummary = cvModel.summary)
from pyspark.ml import Pipeline
from pyspark.ml.classification import LogisticRegression
from pyspark.ml.evaluation import BinaryClassificationEvaluator
from pyspark.ml.feature import HashingTF, Tokenizer
from pyspark.ml.tuning import CrossValidator, ParamGridBuilder
# Prepare training documents, which are labeled.
training = spark.createDataFrame([
(0, "a b c d e spark", 1.0),
(1, "b d", 0.0),
(2, "spark f g h", 1.0),
(3, "hadoop mapreduce", 0.0),
(4, "b spark who", 1.0),
(5, "g d a y", 0.0),
(6, "spark fly", 1.0),
(7, "was mapreduce", 0.0),
(8, "e spark program", 1.0),
(9, "a e c l", 0.0),
(10, "spark compile", 1.0),
(11, "hadoop software", 0.0)
], ["id", "text", "label"])
# Configure an ML pipeline, which consists of tree stages: tokenizer, hashingTF, and lr.
tokenizer = Tokenizer(inputCol="text", outputCol="words")
hashingTF = HashingTF(inputCol=tokenizer.getOutputCol(), outputCol="features")
lr = LogisticRegression(maxIter=10)
pipeline = Pipeline(stages=[tokenizer, hashingTF, lr])
# We now treat the Pipeline as an Estimator, wrapping it in a CrossValidator instance.
# This will allow us to jointly choose parameters for all Pipeline stages.
# A CrossValidator requires an Estimator, a set of Estimator ParamMaps, and an Evaluator.
# We use a ParamGridBuilder to construct a grid of parameters to search over.
# With 3 values for hashingTF.numFeatures and 2 values for lr.regParam,
# this grid will have 3 x 2 = 6 parameter settings for CrossValidator to choose from.
paramGrid = ParamGridBuilder() \
.addGrid(hashingTF.numFeatures, [10, 100, 1000]) \
.addGrid(lr.regParam, [0.1, 0.01]) \
.build()
crossval = CrossValidator(estimator=pipeline,
estimatorParamMaps=paramGrid,
evaluator=BinaryClassificationEvaluator(),
numFolds=2) # use 3+ folds in practice
# Run cross-validation, and choose the best set of parameters.
cvModel = crossval.fit(training)
# Prepare test documents, which are unlabeled.
test = spark.createDataFrame([
(4, "spark i j k"),
(5, "l m n"),
(6, "mapreduce spark"),
(7, "apache hadoop")
], ["id", "text"])
# Make predictions on test documents. cvModel uses the best model found (lrModel).
prediction = cvModel.transform(test)
selected = prediction.select("id", "text", "probability", "prediction")
for row in selected.collect():
print(row)

Answer 1

LogisticRegression model has coefficients not weights. Other than this it can be done as below:
cvModel
# The best model from CrossValidator
.bestModel
# The last stage in Pipeline
.stages[-1]
.coefficients)