Some handy tools for processing essays in a variety of formats and grading them in the command line

A directory containing essays in docx, pages, pdf, and txt formats. Each filename should contain a unique identifier that matches it to a student. A csv file containing names and unique identifiers of students.

Note that the function to read .txt files is specific to their structure in my application and requires modification to read general text files.

[If any of the inputs are .pages files then the script "convert_pages" must be run first to convert them to .pdfs. This will only work on OSX.]

Set path to this directory at top of paper_reader.py to PATH + WEEK. e.g. the input files are in ../data/week1 And set MY_STUDENTS to the path for the csv of student info.

Now in the command line run the script using the command python paper_reader.py The script will then automatically read in the files and create a pandas dataframe containing columns for the student information, the essay text, and the input of a grade.

When the files have been read a command line interface will then print each essay and request you to assign a grade to the essay. A warning will appear if there is no essay available for a given student. A counter will show the number of essays remaining.

For each student simply read the essay and enter the numerical grade.

If you think the text is corrupted or missing then you can assign a missing value for the student and continue grading. The function grade_change can then be used to give these students grades after the problem has been investigated (i.e. manually looking at their submission files)

When all essays have been assigned a grade the dataframe will be saved as a pickle object. This can be unpickled and converted to a csv as necessary.

The function grade_change also allows you to modify the grade of a particular student as necessary. You can load this function in Python to use it.

To view the grade file simply open Python and enter the following code: import pickle \n import pandas as pd \n df = pickle.load(open('weekX_graded.p', 'rb')) #where X is the week number

All essays and final grades can be added into a single dataframe using the table_builder.py file. The INPUT_PATH at the top of the file must be set to the directory containing the papers, where each week's papers are in a directory named by week, e.g. week1, week2, week3, ....

When you run table_builder.py you will be asked to name the file containing all grade information. This is the file that can be downloaded from Blackboard in the Full Grade Center. The code is setup to process grades for ALL students in this file. If you intend to just process one class then you can alter the students variable at the bottom to point to this list of students.

After the code runs a new pickled pandas dataframe will appear in the working directory, where the name is the most recent week followed by _model_table.p.

The required python packages can all be installed using pip, i.e. pip install PyPDF2

Packages: PyPDF2 docx pandas pickle numpy copy string re