analysis.py

# mamba activate jsta
import itertools
import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt


def _state_names():
    # https://gist.github.com/rogerallen/1583593
    us_state_to_abbrev = {
        "Alabama": "AL",
        "Alaska": "AK",
        "Arizona": "AZ",
        "Arkansas": "AR",
        "California": "CA",
        "Colorado": "CO",
        "Connecticut": "CT",
        "Delaware": "DE",
        "Florida": "FL",
        "Georgia": "GA",
        "Hawaii": "HI",
        "Idaho": "ID",
        "Illinois": "IL",
        "Indiana": "IN",
        "Iowa": "IA",
        "Kansas": "KS",
        "Kentucky": "KY",
        "Louisiana": "LA",
        "Maine": "ME",
        "Maryland": "MD",
        "Massachusetts": "MA",
        "Michigan": "MI",
        "Minnesota": "MN",
        "Mississippi": "MS",
        "Missouri": "MO",
        "Montana": "MT",
        "Nebraska": "NE",
        "Nevada": "NV",
        "New Hampshire": "NH",
        "New Jersey": "NJ",
        "New Mexico": "NM",
        "New York": "NY",
        "North Carolina": "NC",
        "North Dakota": "ND",
        "Ohio": "OH",
        "Oklahoma": "OK",
        "Oregon": "OR",
        "Pennsylvania": "PA",
        "Rhode Island": "RI",
        "South Carolina": "SC",
        "South Dakota": "SD",
        "Tennessee": "TN",
        "Texas": "TX",
        "Utah": "UT",
        "Vermont": "VT",
        "Virginia": "VA",
        "Washington": "WA",
        "West Virginia": "WV",
        "Wisconsin": "WI",
        "Wyoming": "WY",
        "District of Columbia": "DC",
        "American Samoa": "AS",
        "Guam": "GU",
        "Northern Mariana Islands": "MP",
        "Puerto Rico": "PR",
        "United States Minor Outlying Islands": "UM",
        "U.S. Virgin Islands": "VI",
    }
    return [x for x in us_state_to_abbrev.keys()]


def grep_name(y, state_names):
    res = [x if x in y else "" for x in state_names]
    if all([len(x) == 0 for x in res]):
        manual_key = {
            "UT Austin": "Texas",
            "UW": "Washington",
            "UPenn": "Pennsylvania",
            "CU Boulder": "Colorado",
        }
        return manual_key[y]

    res = [x for x in itertools.compress(res, [len(x) > 0 for x in res])][0]
    return res


# ---
dt = pd.read_csv("us-faculty-hiring-networks/data/institution-stats.csv")


def lollipop_plot(dt, taxonomy_level="Academia", outpath="figures/lollipop.png"):
    # breakpoint()
    dt = dt[dt["TaxonomyValue"] == taxonomy_level].sort_values("OrdinalPrestigeRank")

    # ---
    state_names = _state_names()
    states = dt[["State" in x for x in dt["InstitutionName"]]].copy()
    states.loc[:, "state"] = [
        states["InstitutionName"].iloc[i].split(" ")[0] for i in range(states.shape[0])
    ]
    states = states[[x in state_names for x in states["state"]]]
    states = states[["InstitutionName", "OrdinalPrestigeRank", "state"]]
    state_names = [x for x in states.state]

    # ---
    unis = dt[["University of" in x for x in dt["InstitutionName"]]].copy()
    unis = unis[[any([x in y for x in state_names]) for y in unis["InstitutionName"]]]
    #
    unis_manual = dt[
        [
            x
            in [
                "UT Austin",
                "Ohio University",
                "UW",
                "UPenn",
                "Indiana University Bloomington",
                "CU Boulder",
            ]
            for x in dt["InstitutionName"]
        ]
    ].copy()
    unis = pd.concat([unis_manual, unis])
    unis.loc[:, "state"] = [grep_name(y, state_names) for y in unis["InstitutionName"]]
    unis = unis.sort_values("OrdinalPrestigeRank")
    unis = unis[
        [x != "Indiana University of Pennsylvania" for x in unis["InstitutionName"]]
    ]
    unis = unis.drop_duplicates("state")
    unis = unis[["InstitutionName", "OrdinalPrestigeRank", "state"]]

    # ---
    # dt[["Colorado" in x for x in dt["InstitutionName"]]].head()

    res = states.merge(unis, how="left", on=["state"])
    # sum(res["OrdinalPrestigeRank_x"] < res["OrdinalPrestigeRank_y"])
    # res[res["OrdinalPrestigeRank_y"] > res["OrdinalPrestigeRank_x"]]
    res["state_minus_uni"] = res["OrdinalPrestigeRank_x"] - res["OrdinalPrestigeRank_y"]
    res["state_minus_uni_abs"] = abs(
        res["OrdinalPrestigeRank_x"] - res["OrdinalPrestigeRank_y"]
    )
    res["min_rank"] = (
        res[["OrdinalPrestigeRank_x", "OrdinalPrestigeRank_y"]]
        .apply(min, axis=1)
        .values
    )
    res = res.sort_values("OrdinalPrestigeRank_y", ascending=False)

    # --- make "lollipop" plot
    viridis = mpl.colormaps["viridis"].resampled(2)
    cmap_hexs = [mpl.colors.rgb2hex(c) for c in viridis.colors]
    conditional_hexs = np.where(
        res["state_minus_uni"] > 0,
        "#ffa600",
        np.where(res["state_minus_uni"] > 0, "#f95d6a", "#a05195"),
    )

    plt.close()
    fig = plt.figure(figsize=(12, 10))
    plt.hlines(
        y=res["state"],
        xmin=res["OrdinalPrestigeRank_x"],
        xmax=res["OrdinalPrestigeRank_y"],
        color=conditional_hexs,
        linewidth=3,
        alpha=0.8,
    )
    # states are x
    plt.scatter(
        res["OrdinalPrestigeRank_x"],
        range(0, len(res)),
        color=conditional_hexs,
        facecolors="none",
        s=80,
        alpha=1,
    )
    # unis are y
    plt.scatter(
        res["OrdinalPrestigeRank_y"],
        range(0, len(res)),
        color=conditional_hexs,
        s=80,
        alpha=1,
    )
    plt.title("Prestige Rank (" + taxonomy_level + ")")
    # plt.show()
    plt.savefig(outpath)


lollipop_plot(dt, "Academia", outpath="figures/lollipop_academia.png")
lollipop_plot(dt, "Natural Sciences", outpath="figures/lollipop_natural.png")