grades.py

from __future__ import print_function, division

import datetime as dt

import pandas as pd
import numpy as np
import scipy.optimize as sopt

def gradebook():
    columns = ("Assignment", "Due", "Grade", "Weight")
    rows = [
        ("Assignment One",                    dt.datetime(2016, 5, 23), np.nan, 0.02),
        ("Assignment Two",                    dt.datetime(2016, 5, 23), np.nan, 0.02),
        ("Peer Feedback One",                 dt.datetime(2016, 5, 30), np.nan, 0.005),
        ("Peer Feedback Two",                 dt.datetime(2016, 5, 30), np.nan, 0.005),
        ("Assignment Three",                  dt.datetime(2016, 5, 30), np.nan, 0.065),
        ("Assignment Four",                   dt.datetime(2016, 5, 30), np.nan, 0.065),
        ("Peer Feedback Three",               dt.datetime(2016, 6, 6),  np.nan, 0.005),
        ("Peer Feedback Four",                dt.datetime(2016, 6, 6),  np.nan, 0.005),
        ("Assignment Five",                   dt.datetime(2016, 6, 6),  np.nan, 0.065),
        ("Peer Feedback Five",                dt.datetime(2016, 6, 13), np.nan, 0.005),
        ("Assignment Six",                    dt.datetime(2016, 6, 13), np.nan, 0.065),
        ("Assignment Seven",                  dt.datetime(2016, 6, 13), np.nan, 0.065),
        ("Peer Feedback Six",                 dt.datetime(2016, 6, 20), np.nan, 0.005),
        ("Peer Feedback Seven",               dt.datetime(2016, 6, 20), np.nan, 0.005),
        ("Assignment Eight",                  dt.datetime(2016, 6, 20), np.nan, 0.065),
        ("Assignment Nine",                   dt.datetime(2016, 6, 20), np.nan, 0.065),
        ("Peer Feedback Eight",               dt.datetime(2016, 6, 27), np.nan, 0.005),
        ("Peer Feedback Nine",                dt.datetime(2016, 6, 27), np.nan, 0.005),
        ("Assignment Ten",                    dt.datetime(2016, 6, 27), np.nan, 0.065),
        ("Peer Feedback Ten",                 dt.datetime(2016, 7, 4),  np.nan, 0.005),
        ("Assignment Eleven",                 dt.datetime(2016, 7, 4),  np.nan, 0.065),
        ("Peer Feedback Eleven",              dt.datetime(2016, 7, 11), np.nan, 0.005),
        ("Term Project Report",               dt.datetime(2016, 7, 25), np.nan, 0.1),
        ("Peer Feedback Term Project Report", dt.datetime(2016, 8, 3),  np.nan, 0.005),
        ("Exam",                              dt.datetime(2016, 8, 3),  np.nan, 0.15),
        ("Final Portfolio",                   dt.datetime(2016, 8, 3),  np.nan, 0.025),
        ("Piazza Participation",              dt.datetime(2016, 8, 3),  np.nan, 0.04),
    ]

    return pd.DataFrame(rows, columns=columns)

def calc(grades):
    df = gradebook()
    df.loc[df['Grade'].isnull(), 'Grade'] = grades
    return np.abs(check_grade(df) - .80)


def optimize(gb=None):
    if gb is None:
        gb = gradebook()
    nullidxs = gb['Grade'].isnull()
    nullsize = gb[nullidxs].shape[0]
    guess = [.5] * nullsize
    bounds = ((0.0, 1.0),) * nullsize
    res = sopt.minimize(
        calc,
        guess,
        method="SLSQP",
        bounds=bounds,
    )
    optimized = gb.copy()
    optimized.loc[nullidxs, 'Grade'] = res.x
    return optimized

def check_grade(df):
    return (df['Grade'] * df['Weight']).sum()

def main():
    result = optimize()
    print(result.to_string())

    print("Optimized for final grade of: {0}".format(check_grade(result)))


if __name__ == "__main__":
    main()