Files

import numpy as np
import pandas as pd
import prairielearn as pl


def generate(data):
    df = pd.read_csv("breast-cancer-train.dat", header=None)

    df2 = pd.DataFrame([
        {
            "city": "Champaign",
            "job": "Professor",
            "age": 35,
            "time": pd.to_datetime("2022-10-06 12:00"),
        },
        {
            "city": "Sunnyvale",
            "job": "Driver",
            "age": 20,
            "time": pd.to_datetime("2020-05-09 12:00"),
        },
        {
            "city": "Mountain View",
            "job": "Data Scientist",
            "age": np.nan,
            "time": pd.to_datetime("2021-12-14 12:00"),
        },
    ])

    dft = pd.DataFrame({
        # Scalars
        "integer": 1,
        "numeric": 3.15,
        "logical": False,
        "character": "foo",
        # Series
        "numeric-list": pd.Series([1.0] * 3).astype("float32"),
        "integer-list": pd.Series([1] * 3, dtype="int8"),
        "character-list": pd.Series(["hello", "world", "stat"]),
        "logical-list": pd.Series([True, False, True]),
        "character-string-list": pd.Series(["a", "b", "c"], dtype="string"),
        # Time Dependency: https://pandas.pydata.org/docs/user_guide/timeseries.html
        "POSIXct-POSIXt-timestamp": pd.Timestamp("20230102"),
        "POSIXct-POSIXt-date_range": pd.date_range("2023", freq="D", periods=3),
        # Categorical: https://pandas.pydata.org/docs/user_guide/categorical.html
        "factor": pd.Categorical(["a", "b", "c"], ordered=False),
        "ordered-factor": pd.Categorical(
            ["a", "b", "c"], categories=["a", "b", "c"], ordered=True
        ),
    })

    data["params"]["df"] = pl.to_json(df.head(15))
    data["params"]["df2"] = pl.to_json(df2, df_encoding_version=2)
    data["params"]["dft"] = pl.to_json(dft, df_encoding_version=2)