The following code provides a streamlined version of the execution of the linear regression model estimation via GLM:

from pyspark.ml import Pipeline

vectorAssembler = feat.VectorAssembler(
    inputCols=forest.columns[1:]
    , outputCol='features')

lr_obj = rg.GeneralizedLinearRegression(
    labelCol='Elevation'
    , maxIter=10
    , regParam=0.01
    , link='identity'
    , linkPredictionCol="p"
)

pip = Pipeline(stages=[vectorAssembler, lr_obj])

(
    pip
    .fit(forest)
    .transform(forest)
    .select('Elevation', 'prediction')
    .show(5)
)