BCS-Python
Using the Wrapper
Installation
The pre-alpha release is available on test.pypi.org. It can be installed just like any other pypi package.
To install
pip install pip install bcs-python==0.0.3
Getting Started
After import, pass in your login email and password to the Bootcampspot constructor.
Note: If you intend to publish anything using this online, anywhere, I would strongly suggest you do not pass the login values into the constructor as shown below. Use a hidden file or, better yet, environmental variables.
>>> from bcs import Bootcampspot
>>> bcs = Bootcampspot(email='johnsmith42@email.com', password='%d-%m-%Y')
>>> bcs.user
{'id': 42,
'userName': 'johnsmith42@email.com',
'firstName': 'John',
'lastName': 'Smith',
'email': 'johnsmith42@email.com',
'githubUserName': 'MyPasswordIsStrfTime',
Use bcs.my_courses for a simple list of Course IDs
>>> bcs.my_courses
[1234,2345]
Use bcs.my_enrollments for the Enrollment IDs
>>> bcs.my_enrollments
[123456]
To get more information about your cohorts, you have a few options:
Simply printing the instance will give you a pretty view (non-interactive consoles)
>>> print(bcs)
| Class Name | courseID | enrollmentId |
|----------------------------------------------------------|
0| UT-MUNICH-UXUI-12-2042-U-C-MW | 1234 | 123456 |
1| UT-MUNICH-UXUI-12-2042-U-C-TTH | 2345 | 123456 |
Or call bcs.class_details for a pandas friendly list of records
>>> bcs.class_details
[{'courseName':'UT-MUNICH-UXUI-12-2042-U-C-MW','courseId':1234, 'enrollmentId': 123456},
{'courseName':'UT-MUNICH-UXUI-12-2042-U-C-TTH','courseId':2345, 'enrollmentId': 234567}]
This is also, temporarily, the output of printing the instance in an interactive terminal, e.g. jupyter notebook
Setting the course will set your enrollment. Your course_id is the most specific identifier for a course so it's a good place to start. By setting the course, you also set the enrollment for your instance.
>>> bcs.course = 1234
>>> bcs.enrollment
123456
If you happen to enter an invalid course/enrollment id, the accompanying error message will give you a list of valid ids to work with. The same is true for setting enrollments.
>>> bcs.course = 1
Traceback (most recent call last):
File "bcswrapper.py", line 399, in <module>
bcs.course = 1
File "bcswrapper.py", line 110, in course
raise CourseError(msg)
CourseError: Invalid courseId '1' not in your courses. Try one of these: [1234, 2345]
Once you've set your bcs.course, both that course and it's accompanying enrollmentId will be used as the default parameters for all method calls. That doesn't mean you can't set your own when you call a method. If you want to make a call for a different course, just enter the corresponding parameter in the method. Your set course will stay set in the background until you manually change it.
Note: If you manually enter an enrollmentId parameter for a cohort with two classes, you will have to also enter a courseId parameter. The code isn't smart enough to know whether you want MW or TTH for a given cohort.
>>> bcs.course = 1234
>>> bcs.grades(courseId=2345)
{'0: Prework':{
'Abe Lincoln': 'A',
'Clifford the Dog': 'Unsubmitted',
'Jane Doe': 'Incomplete'},
'1: Hello, World':{
'Abe Lincoln': 'A+',
'Clifford the Dog': 'B',
'Jane Doe': 'Incomplete'}}
Without that argument the method will default to bcs.course
>>> bcs.grades()
{'0: Prework':{
'Arthur Conan Doyle': 'A',
'Ken Burns': 'Unsubmitted',
'Sterling Archer': 'Incomplete'},
'1: Hello, World':{
'Arthur Conan Doyle': 'A+',
'Ken Burns': 'B',
'Sterling Archer': 'Incomplete'}}
Attendance can be fetched by calling the attendance method. Like the grades method, it will use your set courseId as a default. The resulting data structure (as are all the outputs) is designed to be table friendly so it can be called directly into a DataFrame instance.
>>> import pandas as pd
>>> df = pd.DataFrame(bcs.attendance())
>>> df
Advanced Strings Jiving with JavaScript SQL is Old
Arthur Conan Doyle present present remote
Ken Burns present present present
Sterling Archer remote absent absent
This allows you to do pandas things with your api responses
>>> df['Absences'] = df.isin(['absent']).sum(axis=1)
>>> df
Advanced Strings Jiving with JavaScript SQL is Old Absences
Arthur Conan Doyle present present remote 0
Ken Burns present present present 0
Sterling Archer remote absent absent 2
The method itself sets the boolean values of the API response into categories for a given class to flatten the response. There is some logic applied by the wrapper to decode whatever logic is applied server side before the API call to get more consistent results. I will adjust the code as I understand more about what's going on server side. Just know what the wrapper returns is true so long as your student is actually in that courseId.
@TODO:
- students: info on students for a courseId
- [ x ] weekly_feedback
- assignments
- finish docstrings for sphinx
- setup.py/Manifest.in
- swap over to environment.yml (use this to build requirements-dev.txt)
Known Issues:
Both grades and attendance endpoints return all students from a given enrollment regardless of the chosen course. However, it doesn't appear as though the information from the route (letter grades or attendance values) are reported properly for students in the other courseId. This results in many false negatives for 'absent' and 'not submitted'.
I'm exploring a workaround that will call the session_details uri using the session_closest and populate a list of students for a given courseId and then I'll use the return as a mask for the grades and attendance based on that. It's a hack and requires 3 total api calls instead of the 1 but it seems like the only reliable way to generate a list of students to filter against for as many situations as possible.
