# import modules
import pandas as pd

# import dataframe from .csv
data = pd.read_csv("./data.csv")

# create empty dataframe for output
columnNames = ['user', 'trial', 'reaction_time']
newData = pd.DataFrame(columns=columnNames)

processedTrials = [] # used to mark trials that have been processed
currentUser = 0 # the current user being processed

for index, row in data.iterrows():
    print("Processing row ", index, "...")

    # if processing a new user, clear the list of processed trials and assign new user as current
    if row['user'] != currentUser:
        currentUser = row['user']
        processedTrials = []

    # if selected trial has already been processed, skip
    if row['trial'] in processedTrials:
        print("Skipping row")
        continue
    # otherwise, save data to new dataframe
    else:
        print("Appending data...")
        processedTrials.append(row['trial'])
        newData = newData.append({'user':row['user'], 'trial':row['trial'], 'reaction_time':row['\'reaction_time_2\'']}, ignore_index=True)

# print sample of output and save to .csv
print(newData.head())
newData.to_csv('./newData.csv', index=False, encoding='utf-8')