2_call.py

# Imports
from past.builtins import execfile
execfile('1_imports.py')

# Get parameters given for script execution.
# The 1st parameter is the dataset name to use. It is the name of the file without the extension. The dataset must be a .txt file with ";" as delimiter.
dataset = ""
try:
    dataset = sys.argv[1]
except:
    print("ERROR: You must give a dataset name to use as first parameter. Just give the name without the .txt extension.")
    sys.exit()

# Get Raw Data.
# df must have as last column a categorical variable to predict ("y" variable) (in bmdec it is a combination of transport modes) and as other columns various predictor categorical variables ("x" variables).
df = []
with open("Data/" + dataset + ".txt", newline='') as txtfile:
        df = list(csv.reader(txtfile, delimiter=';'))

# Transform df into a Dataframe.
df = DataFrame(df)
df = df.values

# Scale of the criteria = Scale of df's columns except the first.
# MinMaxScaler() scales values to have them laying between 0 and 1 included.
# scaler = preprocessing.MinMaxScaler()
# df[:,1:] = scaler.fit_transform(df[:,1:])

# Save df as a file.
pd.DataFrame(df).to_csv("Outputs/df.csv", header=None, index=None)

# Parameters to call main.py
# Path to df.csv
path_to_df = "Outputs/df.csv"
# Macro iterations.
macro_iterations = 1
# Algorithms to execute. Also used for files saved.
# Type inside brackets the names of the algorithms you want to execute. 
# Available algorithms : "naive_bayes", "decision_tree", "logistic_regression", "neural_net_categorical", "k_nearest_neighbours", "support_vector_machine", "csp"
# Not suitable for categorical predictions : "linear_regression", "neural_network", "k_means"
algos = ["naive_bayes", "decision_tree", "k_nearest_neighbours", "logistic_regression"]
algo_types = ""
for t in algos:
    algo_types += t + "-"
algo_types = algo_types[:-1] # remove last character ("-")

# Call main.py with its parameters.
#os.system("3_main.py " + path_to_df + " " + str(macro_iterations) + " " + str(algo_types))
subprocess.run(["python", "3_main.py", path_to_df, str(macro_iterations), str(algo_types)])