Student (Thesis) Project Allocation

This project contains:

1. Jupyter Notebook to prepare data for optimising the allocation of students to projects using the optimiser pdn_project_allocation to optimally assign projects to students;
2. Jupyter Notebook to report the automatically obtained Student-Project allocations in Excel spreadsheets suitable for sharing with students and project supervisors.
3. Jupyter Notebook to report the final Student-Project allocations alongside summary statistics in Excel spreadsheets for internal use by the Course Organiser.
4. Jupyter Notebook to optimise Reader allocation (supervisor and 2nd reader) and report the results and statistics.
5. Jupyter Notebook to report the results and statistics of the Marking of the projects (in progress).

01_Allocation-Preparation:

This Jupyter Notebook takes as input:

• Student data spreadsheet containing 3 tabs:
  • Students with the key columns: Full Name and Email Address (which is used to match students to student preferences)
  • Student_preferences (exported from an MS Form),
  • Student_selfproposals (only used for the Allocation reporting below) with the following columns: Student, Email, Code, Supervisor, Title, Subject Area, Type.
• Project data spreadsheet containing 5 tab:
  • Projects with the following columns: Code, Title, Supervisor, Subject Area, Research Type, Level, Max. Students.
  • Subject Areas with the list of allowable Subject Area values.
  • Project Types with the list of allowable Project Type values.
  • Levels with the list of allowable student Level values.
  • Counts with the list of allowable project maximum student Count values.
• Supervisor data spreadsheet containing 2 tabs:
  • Supervisors with the following columns: Supervisor, Supervisor_short, Subject_area, Max_number_of_projects, Max_number_of_students, 2nd_reader.
  • Supervisors_preferences (see pdn_project_allocation for details). The notebook produces the 01_Output.xlsx containing the input data required for running the allocation optimiser pdn_project_allocation.

02_Allocation:

This is done using pdn_project_allocation optimiser project at https://github.com/RudolfCardinal/pdn_project_allocation . You can install it locally using:

pip install git+https://github.com/RudolfCardinal/pdn_project_allocation

NOTE: Note that this package last did not work with Python 3.12 and/or the latest version of the mip library, and instead required Python 3.11 and mip version 1.15.0. This may well change, but see 02_Input_Script.txt for the latest recorded experience and how it was resolved.

The pdn_project_allocation optimiser should be rung as follows (assuming it is run from the the folder containing ./O1_Output.xlsx:

pdn_project_allocation.exe --output .\02_Output.xlsx .\01_Output.xls

Assumptions and Notes:

• Project, Subject Area and Project Type in the Student Preferences spreadsheet are all treated as projects according to the pdn_project_allocation optimiser, but Projects are systematially ranked above Subject Areas, and Subject Areas above Project Types.
• Subject Area for both the projects and supervisors can be list of subject areas. The lists are strings with the subject areas separated by ;# (this is because this is what MS Lists does when exporting into xlsx.
• Project Type is currently discarded because it has not been needed at all for now, but could be treated similarly to Subject Area but as a third tier.
• The Student Preferences selection form asks for top 5 preferred Projects, top 3 preferred Subject Areas, and top 2 preferred Project Types.

03_Allocation-Reporting:

This Jupyter Notebook takes the output spreadsheet of step 2 (./O2_Output_yymmdd-hhmmss.xlsx) as well as the list of Projects and the list of Supervisors, and creates:

• the final project allocation data spreadsheet that is shared with the students and supervisors (./O3_Output_yymmdd-hhmmss.xlsx).
• the final project allocation data and summary statistics spreadsheet for internal use (./O3_Output-Internal_yymmdd-hhmmss.xlsx).

NOTE: If ./O3_Output_yymmdd-hhmmss.xlsx is adjusted manually by the Course Organiser to reflect some request for reallocation, then there is a cell to reload it to recompute ./O3_Output-Internal_yymmdd-hhmmss.xlsx

04_Readers-Allocation:

This Jupyter Notebook takes the list of supervisors and the allocated projects, and then allocate second readers so that the overall load of all supervisors is balanced as much as possible and 2nd readers are allocated projects that fall in their Subject Areas of interest. The Jupyter Notebook outputs both the final reader allocations and supervisor+2nd reader statistics for all people.

05_Marking-Reporting (in progress):

This Jupyter Notebook takes the marking results of the two markers and the agreed marks and report the results and statistics of the marks.

TODO:

• Check one last time that the process for conducting the optimisation and reporting its results works correctly
• Ensure all the files are outputted in one place in a logical naming/order.
• Merge student-related spreadsheet into one spreadsheet (with multiple tabs) and same for supervisor spreadsheet. This way, the input will include just 3 spreadsheets.
• Automate the Reader Allocation process (initiated)
• Merge Readers-Reporting notebook into Readers-Allocation notebook
• Complete the Marking Reporting notebook
• Add Example Data\ folder including all Input spreadsheets.
• (optional) Convert Jupyter Notebooks into simple Python scripts