bacteriaClassification: Classification of bacteria based on their morphological...

In masato-ogishi/GramStainR: Automatic classification of digital Gram stain images

Description

Classification of bacteria based on their morphological features using H2O AutoML instances

Usage

bacteriaClassification_AutoML(trainDF, testDF, destDir = "./Results/",
  LeaderBoardName = "LeaderBoard.csv", H2OModelName = "BestH2OModel",
  seed = 12345, max_mem_size = "6G", nthreads = 6)

bacteriaClassification_Prediction(evalDF, destDir = "./Results/",
  H2OModelName = "BestH2OModel", seed = 12345, max_mem_size = "6G",
  nthreads = 6)

bacteriaClassification_Evaluation(evalDF, targetBacteria = "Paeruginosa",
  destDir = "./Results/", H2OModelName = "BestH2OModel",
  PredictionHeader = "Predictions_", seed = 12345, max_mem_size = "6G",
  nthreads = 6)

Arguments

trainDF	A dataframe for model training.
testDF	A dataframe for hold-out validation.
destDir	A directory for saving outputs.
LeaderBoardName	The file name of the leader board to be saved. The file format must be CSV. If you want to skip saving, set NULL.
H2OModelName	The file name of the best H2O model to be saved. No file extension is required. If you want to skip saving, set NULL.
seed	A random seed.
max_mem_size	The upper limit of memory for H2O Java machine learning engine.
nthreads	The number of threads for H2O Java machine learning engine.
evalDF	A dataframe for evaluation.
targetBacteria	The name of the bacteria of interest.
PredictionHeader	The file name header of the prediction results to be saved.

masato-ogishi/GramStainR documentation built on May 13, 2019, 6:15 p.m.