import pandas as pd
import random
import names
from matplotlib import pyplot as plt

def roundGrade(grades):
    grade_list = [12, 10, 7, 4, 2, 0, -3]
    gradesRounded = []

    for grade in grades:
        gradesRounded.append(min(grade_list, key=lambda x:abs(x-grade)))

    return gradesRounded


def computeFinalGrades(grades):
    data = pd.read_csv(grades)
    min_cols = 3                    # Minimum number of columns for value data

    if data.shape[1] == min_cols:   # Data only has one assignment
        data['Final'] = data.ix[:,2]
    else:                           # Data has two or more assignments
        data['Final'] = roundGrade((data.sum(axis=1) - data.min(axis=1)) /
            (data.shape[1] - min_cols))

    for name in data.ix[:,2:]:      # Student is given final grade -3, if -3 is present
        data.loc[(data[name] == -3), 'Final'] = -3

    gradesFinal = data['Final']

    return gradesFinal


def generate_test_data(assignments, entries, grade_errors):
    grade_list = [12, 10, 7, 4, 2, 0, -3]

    for _ in range(grade_errors):
        grade_list.append(round(random.uniform(-3, 12), 1))

    header = ["StudentID","Name"]

    for n in range(1, assignments + 1):
        header.append(f"Assignment{n}")

    with open('data.csv', 'w') as file:
        file.write(",".join(header))
        file.write("\n")
        for _ in range(entries):
            grades = []
            name = names.get_full_name()
            number = f"s{int(random.uniform(103456, 154321))}"
            file.write(f"{number},{name},")
            for i in range(1, assignments + 1):
                grades.append(str(random.choice(grade_list)))
            file.write(",".join(grades))
            file.write("\n")

generate_test_data(3, 50, 10)

def check_data(filename):
    grade_list = [12, 10, 7, 4, 2, 0, -3]
    df = pd.read_csv(filename)
    assignment_names = df.columns.values[2:]
    grouped_data = df.groupby('StudentID')['StudentID'].agg('count')

    if len(df.index) == len(grouped_data.index):
        for assignment in assignment_names:
            grades = df[assignment_names][df[assignment_names].isin(grade_list)]

        result = pd.concat([df.ix[:,:2], grades], axis=1)
        n_columns = len(result.columns.values)
        result['NaNs'] = n_columns - result.apply(lambda x: x.count(), axis=1)

        if result.isnull().values.any() == True:
            return False, result
        else:
            good_entries = result[result.NaNs <= 0]
            return True, good_entries
    else:
        invalid_ids = []
        for number in grouped_data.index:
            if grouped_data[number] > 1:
                invalid_ids.append(number)

        return False, invalid_ids


data_check, mydata = check_data('data.csv')