Merge pull request #152 from afeld/attendance

add attendance script

Author: afeld

extras/scripts/attendance.py

@@ -0,0 +1,118 @@
+# this is a script rather than a notebook to avoid private student information being committed to the repository
+
+import pandas as pd
+
+NUM_CLASSES = 7
+FREEBIES = 1
+TOP_SCORE = NUM_CLASSES - FREEBIES
+
+ROLL_CALL_CSV = (
+    "~/Downloads/attendance_reports_attendance-264e4d14-1765-4396-b311-4d927b59566d.csv"
+)
+# get by clicking into the Assignment and getting from the URL
+ASSIGNMENT_ID = 1405957
+GRADEBOOK_FILE = "attendance.csv"
+STUDENT_UNIQUE_COLS = ["Student ID", "Student Name", "Section Name", "Section"]
+
+
+def normalize_sections(entries: pd.DataFrame):
+    """For students who switch sections, use their final section."""
+
+    student_info = entries.drop_duplicates(subset=["Student ID"], keep="last")
+    student_info = student_info[["Student ID", "Section Name"]]
+
+    entries = entries.drop(columns=["Section Name"])
+    return entries.merge(student_info, on="Student ID")
+
+
+def get_entries(filename: str):
+    entries = pd.read_csv(
+        filename,
+        index_col=False,
+        usecols=[
+            "Section Name",
+            "Student Name",
+            "Student ID",
+            "Class Date",
+            "Attendance",
+        ],
+        parse_dates=["Class Date"],
+    )
+
+    entries = normalize_sections(entries)
+    # pull the section number out
+    entries["Section"] = (
+        entries["Section Name"].str.extract(r"INAFU6504_(\d{3})_").astype(int)
+    )
+
+    return entries
+
+
+def print_heading(text: str):
+    print(f"-------------------\n\n{text.upper()}:\n")
+
+
+def print_students(students: pd.Series):
+    print(students.droplevel(["Student ID", "Section Name"]))
+    print()
+
+
+def validate(entries: pd.DataFrame):
+    recording_counts = entries.groupby(STUDENT_UNIQUE_COLS).size()
+    print_heading("Students missing entries")
+    print_students(recording_counts[recording_counts < NUM_CLASSES])
+
+    total_classes = entries["Class Date"].nunique()
+    assert total_classes == NUM_CLASSES
+
+
+def compute_scores(entries: pd.DataFrame):
+    attended = entries[entries["Attendance"] == "present"]
+    attendance_counts = attended.groupby(STUDENT_UNIQUE_COLS).size()
+    # print_heading("Attendance counts")
+    # print_students(attendance_counts)
+
+    # cap the top scores
+    attendance_counts[attendance_counts > TOP_SCORE] = TOP_SCORE
+
+    return attendance_counts
+
+
+def write_canvas_csv(scores: pd.Series):
+    """https://community.canvaslms.com/t5/Instructor-Guide/How-do-I-import-grades-in-the-Gradebook/ta-p/807"""
+
+    attendance_col = f"Attendance ({ASSIGNMENT_ID})"
+
+    gradebook = (
+        scores.reset_index(name=attendance_col)
+        .drop(columns=["Section"])
+        .rename(
+            columns={
+                # Roll Call has `FIRST LAST`, gradebook has `LAST, FIRST`. Shouldn't matter.
+                "Student Name": "Student",
+                "Student ID": "ID",
+                "Section Name": "Section",
+            }
+        )
+    )
+    gradebook.to_csv(GRADEBOOK_FILE, index=False)
+
+    print(f"Now upload {GRADEBOOK_FILE} to CourseWorks Gradebook.")
+
+
+def run():
+    entries = get_entries(ROLL_CALL_CSV)
+    validate(entries)
+
+    scores = compute_scores(entries)
+    print_heading("Scores")
+    print_students(scores)
+
+    lowered_scores = scores[scores < TOP_SCORE]
+    print_heading(f"Scores below {TOP_SCORE}")
+    print_students(lowered_scores.sort_values())
+
+    write_canvas_csv(scores)
+
+
+run()

meta/assistant_guide.md

@@ -215,3 +215,14 @@ It isn't your responsibility to look for potential instances of cheating/plagiar
 1. In the Gradebook, give points to the reviewer under the Final Project Peer Review.
 
 [Scoring details.](../syllabus.md#final-project)
+
+{% if id == "columbia" -%}
+## Final grades
+
+To compute the [attendance](../syllabus.md#attendance) score:
+
+1. [Export the Roll Call attendance data.](https://community.canvaslms.com/t5/Canvas-Basics-Guide/What-is-the-Roll-Call-Attendance-Tool/ta-p/59#export_attendance_data)
+1. Copy [the script](https://github.com/afeld/python-public-policy/blob/main/extras/scripts/attendance.py) into {{coding_env_name}}.
+1. Adjust the constants at the top as necessary.
+1. Run the code.
+{%- endif %}