undergrad-uh401/scorecard_tools.py

import os
import logging

import awkward as ak
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd


logger = logging.getLogger("Scorecard Data")


def mask_valid(data):
    # Strip out NaN and zero placeholders
    mask_nan = ~ak.any(np.isnan(data.academics.program_percentage.physical_science), axis=1)
    mask_all_zero = ~ak.all(data.academics.program_percentage.physical_science == 0, axis=1)
    mask = mask_nan & mask_all_zero
    data = data[mask]

    return data

def mask_school(data, name):
    return data[data.school.name[:, -1] == name][0]


# Percentage enrollment per field over time averaged across the nation
def save_scorecard_stats_national(output_dir, data, ext="png"):
    prog_percentage = data.academics.program_percentage
    students = data.student.enrollment.undergrad_12_month

    # Percent enrollment
    for key in prog_percentage.fields:
        # Find the mean across the nation
        mean = ak.mean(prog_percentage[key], axis=0)
        if ak.all(mean < 0.05):
            continue
        plt.plot(mean * 100, label=key)
    plt.xlabel("Year")
    plt.ylabel("Percent Enrollment")
    plt.title("Average Enrollment (National)")
    plt.legend()
    plt.savefig(os.path.join(output_dir, f"national-enrollment-percent.{ext}"))
    plt.clf()

    # Students enrolled
    for key in prog_percentage.fields:
        # Find the mean across the nation
        mean_perc = ak.mean(prog_percentage[key], axis=0)
        if ak.all(mean_perc < 0.05):
            continue
            
        mean = ak.mean(prog_percentage[key] * students, axis=0)
        plt.plot(mean, label=key)
    plt.xlabel("Year")
    plt.ylabel("Percent Enrollment")
    plt.title("Average Enrollment (National)")
    plt.legend()
    plt.savefig(os.path.join(output_dir, f"national-enrollment-students.{ext}"))
    plt.clf()

    
# Percentage enrollment per field over time averaged across one school
def save_scorecard_stats_school(output_dir, data, school: str, ext="png"):
    data = mask_school(data, school)
    prog_percentage = data.academics.program_percentage
    students = data.student.enrollment.undergrad_12_month

    # Plot percent enrollment
    for key in prog_percentage.fields:
        # Find the mean across the nation
        mean = ak.mean(prog_percentage[key], axis=0)
        if mean < 0.05:
            continue
        plt.plot(prog_percentage[key] * 100, label=key)
    plt.xlabel("Year")
    plt.ylabel("Percent Enrollment")
    plt.title(f"Average Enrollment ({school})")
    plt.legend()
    plt.savefig(os.path.join(output_dir, f"{school}-enrollment-percent.{ext}"))
    plt.clf()

    # Plot student enrollment
    for key in prog_percentage.fields:
        # Find the mean across the nation
        mean = ak.mean(prog_percentage[key], axis=0)
        if mean < 0.05:
            continue
        plt.plot(prog_percentage[key] * students, label=key)
    plt.xlabel("Year")
    plt.ylabel("Percent Enrollment")
    plt.title(f"Average Enrollment ({school})")
    plt.legend()
    plt.savefig(os.path.join(output_dir, f"{school}-enrollment-students.{ext}"))
    plt.clf()

def save_scorecard_stats(output_dir, data, ext="png"):
    # Students Enrolled Distribution (Average)
    avg_students = ak.mean(data.student.enrollment.undergrad_12_month, axis=1)
    plt.hist(avg_students, bins=np.arange(0, 100_000, 1_000))
    plt.xlabel("Students Enrolled")
    plt.ylabel("Univeristy Count")
    plt.title("Enrollemnt Distribution (Average)")
    plt.yscale("log")
    plt.savefig(os.path.join(output_dir, f"dist-students-enrolled-avg.{ext}"))
    plt.clf()

    # Students Enrolled Distribution (Last Year)
    plt.hist(data.student.enrollment.undergrad_12_month[:, -1], bins=np.arange(0, 100_000, 1_000))
    plt.xlabel("Students Enrolled")
    plt.ylabel("Univeristy Count")
    plt.title("Enrollemnt Distribution (Last Year)")
    plt.yscale("log")
    plt.savefig(os.path.join(output_dir, f"dist-students-enrolled-last-year.{ext}"))
    plt.clf()



# Print national enrollment stats
def print_enrollment_national(data):
    prog_percentage = data.academics.program_percentage
    students = data.student.enrollment.undergrad_12_month
    
    # Create dataframe of enrollment across the nation (latest academic year and average)
    df = {'Name': list(prog_percentage.fields), 'Percent (Last)': [], 'Percent (Avg)': [], "Students (Last)": [], "Students (Avg)": [] }
    for key in df['Name']:
        vals = prog_percentage[key]
        df['Percent (Last)'].append(ak.mean(vals[-1]) * 100)
        df['Percent (Avg)'].append(ak.mean(vals) * 100)
        df['Students (Last)'].append(ak.mean(vals[-1] * students[-1]))
        df['Students (Avg)'].append(ak.mean(vals * students))
        
    df = pd.DataFrame(df)

    # Print Dataframe
    logger.info("\n\nNational Percentage (Last Academic Year): ")
    print(df.sort_values("Percent (Last)", ascending=False))
    
    logger.info("\n\nNational Percentage (Avg): ")
    print(df.sort_values("Percent (Avg)", ascending=False))

def print_enrollment(data, name: str):
    data = mask_school(data, name)
    prog_percentage = data.academics.program_percentage
    students = data.student.enrollment.undergrad_12_month
    
    # Create dataframe of enrollment across the nation (latest academic year and average)
    df = {'Name': list(prog_percentage.fields), 'Percent (Last)': [], 'Percent (Avg)': [], "Students (Last)": [], "Students (Avg)": [] }
    for key in df['Name']:
        vals = prog_percentage[key]
        df['Percent (Last)'].append(vals[-1] * 100)
        df['Percent (Avg)'].append(ak.mean(vals) * 100)
        df['Students (Last)'].append(vals[-1] * students[-1])
        df['Students (Avg)'].append(ak.mean(vals * students))
        
    df = pd.DataFrame(df)

    # Print Dataframe
    logger.info(f"\n\nProgram Percentage ({name}, Last Academic Year): ")
    print(df.sort_values("Percent (Last)", ascending=False))
    
    logger.info(f"\n\nProgram Percentage ({name}, Avg): ")
    print(df.sort_values("Percent (Avg)", ascending=False))