tests/test_metrics.py - master - nevers/hackINTERPOL

tests/test_metrics.py
81 строка3 KB
nevers
upload files
13 янв 2026, 17:05
13 янв 2026, 17:05949f79e
"""Tests for metrics computation."""

import pandas as pd
import pytest

from studypulse.metrics import (
    compute_descriptive_stats,
    compute_risk_flags,
    compute_risk_score,
    get_top_risks,
)


@pytest.fixture
def sample_df() -> pd.DataFrame:
    """Create sample DataFrame for testing."""
    return pd.DataFrame(
        {
            "student_id": ["STU001", "STU002", "STU003"],
            "student_name": ["Alice", "Bob", "Charlie"],
            "assignment_score": [85.0, 45.0, 95.0],
            "assignments_submitted": [8, 4, 9],
            "attendance_rate": [0.9, 0.6, 0.98],
            "days_since_last_submission": [3, 18, 1],
            "days_to_deadline": [5, -3, 7],
        }
    )


def test_compute_descriptive_stats(sample_df: pd.DataFrame) -> None:
    """Test descriptive stats computation."""
    stats = compute_descriptive_stats(sample_df)
    assert "total_students" in stats
    assert stats["total_students"] == 3
    assert "mean_assignment_score" in stats
    assert "mean_attendance_rate" in stats
    assert "low_score_count" in stats


def test_compute_risk_flags(sample_df: pd.DataFrame) -> None:
    """Test risk flag computation."""
    df_with_flags = compute_risk_flags(sample_df)
    assert "risk_flag" in df_with_flags.columns
    # STU002 should be flagged (low score, low attendance, inactive, behind deadline)
    assert df_with_flags.loc[df_with_flags["student_id"] == "STU002", "risk_flag"].iloc[0] == 1
    # STU003 should not be flagged (high score, good attendance)
    assert df_with_flags.loc[df_with_flags["student_id"] == "STU003", "risk_flag"].iloc[0] == 0


def test_compute_risk_score(sample_df: pd.DataFrame) -> None:
    """Test risk score computation."""
    df_with_score = compute_risk_score(sample_df)
    assert "risk_score" in df_with_score.columns
    # Risk scores should be between 0 and 100
    assert df_with_score["risk_score"].min() >= 0
    assert df_with_score["risk_score"].max() <= 100
    # Higher risk students should have higher scores
    assert (
        df_with_score.loc[df_with_score["student_id"] == "STU002", "risk_score"].iloc[0]
        > df_with_score.loc[df_with_score["student_id"] == "STU003", "risk_score"].iloc[0]
    )


def test_get_top_risks(sample_df: pd.DataFrame) -> None:
    """Test top risks retrieval."""
    df_with_score = compute_risk_score(sample_df)
    top_risks = get_top_risks(df_with_score, top_n=2)
    assert len(top_risks) == 2
    assert "risk_score" in top_risks.columns
    # Should be sorted descending by risk_score
    assert top_risks["risk_score"].iloc[0] >= top_risks["risk_score"].iloc[1]
    # STU002 should be in top risks
    assert "STU002" in top_risks["student_id"].values


def test_metrics_stable_ordering(sample_df: pd.DataFrame) -> None:
    """Test that metrics return stable ordering."""
    df_with_score1 = compute_risk_score(sample_df)
    df_with_score2 = compute_risk_score(sample_df)
    # Risk scores should be deterministic
    pd.testing.assert_frame_equal(df_with_score1, df_with_score2)