ml-regression-logistic.md
In pysparklyr: Provides a 'PySpark' Back-End for the 'sparklyr' Package

Logistic Regression works with Spark Connection

Code
  class(ml_logistic_regression(sc, max_iter = 10))
Output
  [1] "ml_connect_estimator"        "ml_logistic_regression"     
  [3] "ml_probabilistic_classifier" "ml_classifier"              
  [5] "ml_predictor"                "ml_estimator"               
  [7] "ml_pipeline_stage"

Logistic Regression works with `tbl_spark`

Code
  model$features
Output
   [1] "mpg"  "cyl"  "disp" "hp"   "drat" "wt"   "qsec" "vs"   "gear" "carb"

Code
  model$label
Output
  [1] "am"

Code
  model %>% ml_predict(tbl_mtcars) %>% colnames()
Output
   [1] "mpg"         "cyl"         "disp"        "hp"          "drat"       
   [6] "wt"          "qsec"        "vs"          "am"          "gear"       
  [11] "carb"        "prediction"  "probability"

Logistic Regression works with Pipeline

Code
  cap_out[c(1, 3:4, 6:18)]
Output
   [1] "Pipeline (Estimator) with 1 stage"    
   [2] "  Stages "                            
   [3] "  |--1 LogisticRegression (Estimator)"
   [4] "  |    (Parameters)"                  
   [5] "  |    batchSize: 32"                 
   [6] "  |    featuresCol: features"         
   [7] "  |    fitIntercept: TRUE"            
   [8] "  |    labelCol: label"               
   [9] "  |    learningRate: 0.001"           
  [10] "  |    maxIter: 100"                  
  [11] "  |    momentum: 0.9"                 
  [12] "  |    numTrainWorkers: 1"            
  [13] "  |    predictionCol: prediction"     
  [14] "  |    probabilityCol: probability"   
  [15] "  |    seed: 0"                       
  [16] "  |    tol: 1e-06"