Question

Python noob here. I have the following in my database in an object called “data” and I want to write some python that will select the lowest score from each group of student_id….and grab the unique ObjectId for each minimum record found?

{u'student_id': 197, u'_id': ObjectId('50906d7fa3c412bb040eb88d'), u'type': u'homework', u'score': 10.90872422518918}
{u'student_id': 197, u'_id': ObjectId('50906d7fa3c412bb040eb88e'), u'type': u'homework', u'score': 88.3871242475841}
{u'student_id': 198, u'_id': ObjectId('50906d7fa3c412bb040eb892'), u'type': u'homework', u'score': 17.46279901047208}
{u'student_id': 198, u'_id': ObjectId('50906d7fa3c412bb040eb891'), u'type': u'homework', u'score': 76.18366499496366}
{u'student_id': 199, u'_id': ObjectId('50906d7fa3c412bb040eb895'), u'type': u'homework', u'score': 49.34223066136407}
{u'student_id': 199, u'_id': ObjectId('50906d7fa3c412bb040eb896'), u'type': u'homework', u'score': 58.09608083191365}

Thank you for your trouble….here is my code…

import pymongo

from itertools import groupby
from pymongo import Connection


connection = Connection('localhost', 27017)

db = connection.students

data = db.grades.find({'type' : 'homework'}).sort([('student_id',pymongo.ASCENDING),('score',pymongo.ASCENDING)])



for student_id, items in groupby(data, lambda s: s['student_id']):
    lowest_score = min(i['score'] for i in items)

    print lowest_score

Answer 1

I think you may be able to do your minimum check differently and get the ObjectID rather than the score from the minimum value. Here’s how I’d do it:

lowest_id = min(items, key=lambda i: i['score'])['_id']

Now, it’s possible (indeed, likely) that you can do something similar in the database itself, and probably that will have better performance. However, the above may work just fine if your data set is not too unwieldy.