ub-grader 0.3.5

Librería para auto-corrección local de prácticas con reporte cifrado y firmado

ub-grader

Python library for students to self-grade assignments locally and generate reports to submit to instructors.

Table of Contents

1. Quick Installation
2. Student Usage
3. Spec Format
4. Troubleshooting
5. License

1. Quick Installation

pip install ub-grader

Requires Python 3.10+.

2. Student Usage

Basic Usage

from ub_grader import init_students, load_spec, grade
from solution import solve  # your target function

# Register students (single student or multiple for group work)
init_students([
  {"niub": "A123", "first_name": "Alice", "last_name": "Doe"},
])

# Load assignment specification
load_spec("https://server/assignments/p1.json")  # or file:///abs/path/p1.json

# Grade your solution
result = grade(
  solve,
  student_id="A123",
)

print("Final score:", result["final_score"], "/", result["max_score"])

After running, a report file will be generated. Submit this file to your instructor.

Step-by-Step Guide

1. Install the library: pip install ub-grader
2. Create your solution: Write your solution function in a Python file
3. Register: Use init_students() with your student ID and name (supports single or multiple students)
4. Load spec: Use load_spec() with the assignment specification URL provided by your instructor
5. Grade: Use grade() with your solution function
6. Submit: Send the generated report file to your instructor

Multiple Students (Group Work)

You can register multiple students at once, which is useful for group assignments or when managing multiple student accounts:

init_students([
    {"niub": "A123", "first_name": "Alice", "last_name": "Doe"},
    {"niub": "B456", "first_name": "Bob", "last_name": "Smith"},
    {"niub": "C789", "first_name": "Charlie", "last_name": "Brown"},
])

# Each student can then generate their own report
result_alice = grade(solve, student_id="A123")
result_bob = grade(solve, student_id="B456")

3. Spec Format

Assignment specifications are provided by your instructor as JSON files. A typical spec includes:

{
  "version": "1.0.0",
  "assignment_id": "p1",
  "tests": [
    {
      "id": "t1",
      "input": { "args": [1, 2], "kwargs": {} },
      "expected": 3,
      "weight": 1,
      "expected_hidden": false,
      "time_limit_ms": 500,
      "memory_limit_kb": 10000,
      "comparison": "equal"
    }
  ],
  "scoring": {
    "mode": "weighted_sum_with_penalties",
    "rounding": 2,
    "penalties": {},
    "max_score": 10
  }
}

Key points for students:

• Tests may have expected_hidden: true, which means you won't see the expected output, only pass/fail
• Each test has time and memory limits
• Exceeding limits may result in penalties applied to your score

4. Troubleshooting

Common Issues

Issue	Likely Cause	Solution
RuntimeError: No spec loaded	Forgot to call load_spec()	Call load_spec() before grade()
ValueError: Missing required field	Malformed spec JSON	Contact your instructor for a valid spec
Very low score	Tests failing or penalties applied	Review your logic and check time/memory limits

Best Practices

• Use a dedicated virtual environment: python -m venv myenv && source myenv/bin/activate
• Don't modify the assignment spec file
• Test your solution thoroughly before final grading
• Keep backups of your solution code

Getting Help

• Check that your Python version is 3.10 or higher
• Ensure you have the latest version: pip install -U ub-grader
• Review the error messages carefully - they usually indicate the specific issue
• Contact your instructor if you suspect issues with the assignment specification

5. License

MIT