run_classification.py

import constants as cst
import subprocess as sp
import csv
import numpy as np
import re
import os

import signal
import sys
import util


def run_random_forest1(base_dir_name, experiment_name):
    file_path = cst.BASE + base_dir_name + "/" + experiment_name + "/out.arff"
    if not os.path.exists(file_path):
        print('no file', file_path)
        return None
    try:
        output = str(sp.check_output(["java", "-cp", cst.WEKA_BIN,
                                      "weka.classifiers.trees.RandomForest", '-P', '100', '-I',
                                      '100', '-num-slots', '1', '-K', '0', '-M', '1.0', '-V', '0.001', '-S', '1',
                                      "-t", file_path], timeout=30))
        match = re.search('Correctly Classified Instances\s+\d+\s+(\d+\.?\d+)\s+%', output)
        if match:
            return experiment_name + "," + match.group(1)
        return experiment_name + ",?"
    except sp.TimeoutExpired:
        print("timeout exceeded", experiment_name)
        return experiment_name + ",?"


def classify_experiments(base_dir_name, class_arr):
    global results
    results = []
    listdir = os.listdir(cst.BASE + base_dir_name)
    # listdir = None
    # with open(cst.BASE + base_dir_name + "/experiments_list.txt", 'r') as f:
    #     listdir = f.readlines()
    # # listdir = os.listdir(cst.BASE + base_dir_name)
    # if listdir is None or len(listdir) == 0:
    #     return results
    total = len(listdir)
    counter = 0
    for experiment in listdir:
        counter += 1
        if class_arr is not None and experiment in class_arr:
            print(experiment, "has already been processed", counter, "/", total)
            continue
        # if experiment.startswith("xor"):
        #     continue
        print('running random forest on', experiment, counter, "/", total)
        classification = run_random_forest1(base_dir_name, experiment.strip())
        if not classification:
            continue
        results.append(classification)
        results.append("\n")
    return results

def signal_handler(signal, frame):
    global stop_signal, results
    stop_signal = True
    print('Writing down Classification.csv')
    with open(cst.BASE + base_dir_name + "/Classification.csv", "a") as f:
        f.writelines(results)
    sys.exit(0)

global stop_signal
stop_signal = False
if __name__ == '__main__':
    base_dir_name = sys.argv[1]
    signal.signal(signal.SIGINT, signal_handler)
    class_arr = util.read_csv(cst.BASE + base_dir_name + "/Classification.csv")
    res = classify_experiments(base_dir_name, class_arr)
    with open(cst.BASE + base_dir_name + "/Classification.csv", "a") as f:
        f.writelines(res)