Canvas, grade by question

 Most of the questions are multiple-choice, etc., that are scored automatically by Canvas.  You will need to manually grade question 7, where they are asked to write some code.

 

The rubric for this question is embedded in the instructor’s answer (“General Answer Comments”).  

 

 

Recommended grading procedure:

 

1. Enter SpeedGrader for the quiz. 
2. Up at the top, click the Settings “gear”. Under Options, select “grade by question”.  Save those settings.   This lets you grade question by question rather than person by person. You will get a little selection box near the top:   Choose the question you want to grade and as you move from student to student, it will immediately focus on that question.
1. Move from student to student using the left/right arrows at the top of the right column.
2. There’s still an issue with the fact that each student gets a randomly selected question from the pool. So there may be 4-6 different “question 2”s. Usually I make multiple passes over the students, grading a distinct question each time. This is particularly true with questions where the students are typing proofs, because it’s hard for me as a grader to keep switching context from one proof problem to the next.

 

When you have graded the entire quiz,

1. Edit the quiz settings to Let Students See the Correct Answers
2. On the Grades sheet, Post the grades.

 

 Don’t forget that you will have two versions of the exam, the regular one and the lockdown browser version to go through.

 

UTC Canvas sandbox request

 

 If you ever need another sandbox, you can let us know using this form: https://forms.office.com/r/95Zs91FHCQ

 

Canvas rearrange grade columns

 

Total grades can be moved to the front of the Grade book. 

Canvas gradebook, weighted final grade

Under "Assignments", "Groups" can be added, modified, deleted.

For weighted final grades,
"Assignment" -> "..." -> "weight final grade"

*** Socrative responses to Canvas grade book

Save all socrative report as xlsx files.

Rename all report files using the dates.

Using a Jupyter notebook to analyze NetworkdID in each file, summarize them in Pandas dataframe, output to Excel file.

Create assignment on Canvas for attendance by dates. (This step was skipped on 20190220).
Then Export gradebook from Canvas as csv.

Update the grade csv files. (Make sure the point row was preserved. This can be done to sort the grade on a separate sheet, and then copy the sorted and merged grade back).

Upload grade csv files to Canvas gradebook. Visually check some changes. Wait for grades change to be updated.

In [1]:

import os
import pandas as pd
debug = 1

load class roster from gradebook

In [2]:

mygradebook = '../grades/2019-01-26T2215_Grades-Data_Structures_and_Program_Design.csv'

In [3]:

import csv
#with open(mygradebook, newline='') as csvfile:
#    grades = csv.reader(csvfile, delimiter=' ', quotechar='|')
csvfile = open(mygradebook, newline='')
grades = csv.reader(csvfile)
name2id = {}
id2name = {}
for row in grades:
        print(row[0] +  "\t\t" + row[2])
        if (row[0] != "Student") & (row[0] != "Test Student") & (row[0] !="    Points Possible"):
            name2id[row[0]] = row[2]
            id2name[row[2]] = row[0]

In [5]:

id2name

Out[5]:

In [6]:

grades.line_num

Out[6]:

27

In [7]:

len(id2name)

Out[7]:

24

In [8]:

id2attendance = {} #attendance dictionary

In [9]:

report_df = pd.DataFrame(pd.Series(id2name), columns=["name"])
report_df

Out[9]:

read socrative reports

In [10]:

os.chdir("/Users/hqin/github/socrative/Socrative")
mypath = 'cpsc1110Sp2019_cleaned'
filenames = os.listdir(mypath) # returns list
filenames.sort()
print(filenames)

['01_08_2019.xlsx', '01_10_2019.xlsx', '01_15_2019.xlsx', '01_17_2019.xlsx', '01_22_2019.xlsx', '01_24_2019.xlsx']

In [11]:

from openpyxl import load_workbook
for i in range(len(filenames)): 
#for i in range(0,1): 
    myfilename = filenames[i]
    mycolumnlabel = myfilename.replace(".xlsx","")
    if (debug > 0):
                print("i=" , i , "myfilename =" , myfilename )
    wb = load_workbook(filename = mypath + "/" + myfilename)
    ws = wb["Sheet1"]

    unqiue_netids = set()
    for i in range(1,50):
        current_value = ws.cell(row=i, column=5).value
        if type(current_value) == type("x"):
            current_value = current_value.upper().replace(' ','') #remove white spaces
            #if (debug > 0):
                #print( current_value )
            unqiue_netids.add(current_value)
    
    for id in id2name: 
        if id in unqiue_netids: 
            id2attendance[ id ] =  1
        else:
            id2attendance[ id ] =  0   
    
    print(id2attendance)

    report_df[mycolumnlabel] = pd.Series(id2attendance,index=id2attendance.keys() )

In [12]:

report_df

Out[12]:

In [13]:

report_df.to_excel("socrative_summary.xlsx", sheet_name='scorative_report', index_label ="netid")  # doctest: +SKIP

In [ ]:

Canvas FAQ, tips

-> disable color overlay for course images
-> setting for import
-> current UTC Canvas does not support pick questions based types from test bank. To work around, I can use blackboard to pick question types, export them, and import to Canvas.