def process_forum_entries():
"""
input file: "forum_entries_courses.csv"
header: service,course_id,kurs,semester,description,user,name,nid,id,parent_id,date,subject_length,text_length
returns a dict with:
(kurs,semester) -> "username" -> (written count [int],
avg subject length [float],
median text length [int])
"""
KURS_COL = 2
SEMESTER_COL = 3
USER_COL = 5
SUBJ_COL = 11
TEXT_COL = 12
def handle_forum_entries(grouped):
result = {}
for username in grouped:
if not is_valid_matrikel_nummer(username):
continue
value = grouped[username]
written_count = len(value)
subject_lengths = [int(line[SUBJ_COL]) for line in value]
text_lengths = [int(line[TEXT_COL]) for line in value]
result[username] = (written_count, average(subject_lengths), median(text_lengths))
return result
lines = load_lines("forum_entries_courses.csv")
by_kurs_semester = groupby(lines, [KURS_COL, SEMESTER_COL])
result = groupby(by_kurs_semester, [USER_COL], handle_forum_entries)
return result
def process_entries_read(kurs_mapping):
"""
input:
kurs_mapping - dict with: forum_course_id -> (kurs,semester)
input file: "forum_readlist.py"
header: service,course_id,username,nid,id
returns a dict with
(kurs,semester) -> "username" -> read count [integer]
"""
COURSE_ID_COL = 1
USERNAME_COL = 2
def calculate_read(grouped):
return dict((username, len(grouped[username])) for username in grouped\
if is_valid_matrikel_nummer(username))
lines = load_lines("forum_readlist.csv")
by_course_id = groupby(lines, [COURSE_ID_COL])
by_user = groupby(by_course_id, [USERNAME_COL], calculate_read)
result = {}
for course_id in by_course_id:
key = kurs_mapping[course_id]
result[key] = by_user[course_id]
return result
def process_median_feedback_length():
"""
input file: "abgabe_feedback_courses.csv"
header: service,course_id,kurs,semester,description,user_id,task,subtask,author,comment_length
returns a dict with
(kurs,semester) -> "username" -> median feedback [float]
"""
KURS_COL = 2
SEMESTER_COL = 3
USER_COL = 5
COMMENT_LENGTH_COL = 9
def calculate_feedback(grouped):
result = {}
for username in grouped:
if not is_valid_matrikel_nummer(username):
continue
value = grouped[username]
feedback_lengths = [int(line[COMMENT_LENGTH_COL]) for line in value]
median_feedback = median(feedback_lengths)
result[username] = median_feedback
return result
lines = load_lines("abgabe_feedback_courses.csv")
by_kurs_semester = groupby(lines, [KURS_COL, SEMESTER_COL])
result = groupby(by_kurs_semester, [USER_COL], calculate_feedback)
return result
def process_avg_score():
"""
input file: "abgabe_assessment_results_courses.csv"
header: service,course_id,kurs,semester,description,user_id,result_id,result_value
returns a dict with
(kurs,semester) -> "username" -> (quality score [float],
avg score [float],
number of scores [int])
where quality score:
avg score * number of scores
"""
KURS_COL = 2
SEMESTER_COL = 3
USER_COL = 5
RESULT_ID_COL = 6
RESULT_SCORE_COL = 7
def calculate_avg_score(grouped):
result = {}
for username in grouped:
if not is_valid_matrikel_nummer(username):
continue
value = grouped[username]
by_result_id = groupby(value, [RESULT_ID_COL])
result_scores = [float(line[0][RESULT_SCORE_COL]) \
for line in by_result_id.values()]
avg_score = average(result_scores)
num_scores = len(by_result_id.keys())
max_score = max(result_scores)
if max_score == 0:
quality_score = 0
else:
quality_score = float(sum(result_scores)) / float(max_score)
result[username] = (quality_score, avg_score, num_scores)
return result
lines = load_lines("abgabe_assessment_results_courses.csv")
by_kurs_semester = groupby(lines, [KURS_COL, SEMESTER_COL])
result = groupby(by_kurs_semester, [USER_COL], calculate_avg_score)
return result
def process_plus_count():
"""
input file: "abgabe_assessment_pluses_courses.csv"
header: service,course_id,kurs,semester,description,user_id,plus_date
returns a dict with
(kurs,semester) -> "username" -> plus count [integer]
"""
def count_pluses(d):
return dict((x, len(d[x])) for x in d\
if is_valid_matrikel_nummer(x))
lines = load_lines("abgabe_assessment_pluses_courses.csv")
KURS_COL = 2
SEMESTER_COL = 3
USER_COL = 5
by_kurs_semester = groupby(lines, [KURS_COL, SEMESTER_COL])
result = groupby(by_kurs_semester, [USER_COL], count_pluses)
return result
def process_kurs_mapping():
"""
input file: "forum_entries_courses.csv"
header: service,course_id,kurs,semester,description,user,name,nid,id,parent_id,date,subject_length,text_length
returns a dict with:
(course_id) -> (kurs,semester)
"""
KURS_COL = 2
SEMESTER_COL = 3
COURSE_ID_COL = 1
lines = load_lines("forum_entries_courses.csv")
by_kurs_semester = groupby(lines, [KURS_COL, SEMESTER_COL])
return dict((by_kurs_semester[x][0][COURSE_ID_COL], x)\
for x in by_kurs_semester)
def calculate_avg_score(grouped):
result = {}
for username in grouped:
if not is_valid_matrikel_nummer(username):
continue
value = grouped[username]
by_result_id = groupby(value, [RESULT_ID_COL])
result_scores = [float(line[0][RESULT_SCORE_COL]) \
for line in by_result_id.values()]
avg_score = average(result_scores)
num_scores = len(by_result_id.keys())
max_score = max(result_scores)
if max_score == 0:
quality_score = 0
else:
quality_score = float(sum(result_scores)) / float(max_score)
result[username] = (quality_score, avg_score, num_scores)
return result
from common import transpose, first, groupby, cat
import unittest
if __name__ == '__main__':
assert tuple(transpose(((1, 2, 3), (4, 5, 6)))) == ((1, 4), (2, 5), (3, 6))
assert first('abc') == first(['a', 'b', 'c']) == 'a'
assert cat(['a', 'b', 'c']) == 'abc'
assert (groupby(['test', 'one', 'two', 'three', 'four'], key=len) == {
3: ['one', 'two'],
4: ['test', 'four'],
5: ['three']
})
unittest.main()
