Scheduler

This program is designed to facilitate daily shift scheduling - it was originally designed to facilitate on call rotation scheduling for JHU ResLife. If you want to use it for some other purpose, go ahead! Feel free to use and modify this program as you see fit.

Online

The online version of the scheduler is located at: https://hsghori.github.io/scheduler/.

Development

This web-application is relatively simple and ports the algorithms defined in the schedule.py file to a single page web app that could be hosted for free on github pages. For that reason I did not use any web frameworks like Django / Flask or JS frameworks like React or Angular to clean up the front end logic. The frontend logic and scheduling algorithms logic are written entirely in Javascript using JQuery and the computational load is handled by the client's browser.

You need to have NodeJS installed to properly build this app.

Development dependencies:

• Styledot
• Lodash
• JQuery
• Browserify
• Sass

These dependencies are defined in the package.json file and can be installed directly from there.

DO NOT MODIFY bundle.js or main.css. If you want to make js changes you should update main.js, schedule.js, or constants.js. If you want to make CSS changes you should update base.scss.

When you make any CSS or JS changes you need to run make at the root of the project.

If you want to update the scheduling algorithm I would recommend updating and testing with the python file first before updating the algorithm in JS.

Local

Installation

To use this program you need to have Python 2.7+ installed: https://www.python.org/downloads/. You also need to clone or download this repo. All the requirements are defined in the Pipfile - you can install the dependencies using:

$ pipenv install

Creating a Schedule

1. Create an input text file with the format

   person1 | day_of_week1, day_of_week2, ... | date1, date2, ...
   

   day_of_weekN is the day of the week when person1 can't do duty (ie monday, tuesday, etc) dateN is a specific date where person1 can't do duty (format MM/DD/YYYY) (ie 01/10/2018)

   Example:

   John | Monday | 02/14/2018, 03/01/2018, 05/01/2018
   Jill | Tuesday, Wednesday | 04/10/2018, 05/13/2018
   

2. Save the input file as a text file (someFile.txt) in the same place that you've cloned / downloaded this repo.

3. Open your terminal and navigate to the place where you've cloned / downloaded this repo and type:

   $ python scheduler.py --infile someFile.txt

   The program should output the total number of weekday and weekend assignments for each person and any days where a conflict couldn't be easily resolved. You can also specify the name of the output file (by default is schedule_out.txt) via:

   $ python scheduler.py --infile someFile.txt --outfile someOutput.txt

   If you're using this tool past Spring 2018 you can enter new start and end dates both for the semester and for major breaks (Thanksgiving / Spring)

   For example, the command to generate a schedule for Fall 2018 may look like:

   $ python scheduler.py --infile someFile.txt --outfile someOutput.txt --start-date 8/25/2018 --end-date 12/21/2018 --break-start-date 11/17/2018 --break-end-date 11/25/2018

   If you need help with the commands you can type:

   $ python scheduler.py --help

   Note: Each of the commands above has a less verbose version. In order they are:

   $ python scheduler.py -i someFile.txt
   $ python scheduler.py -i someFile.txt -o someOutput.txt
   $ python scheduler.py -i someFile.txt -o someOutput.txt -s 8/25/2018 -e 12/21/2018 -bs 11/17/2018 -be 11/25/2018
   $ python scheduler.py -h

4. If the program ran with no errors a file (someOutput.txt) should have been generated with a randomized duty schedule based on the restrictions given in someFile.

Note: This program isn't perfect. It uses a lot of randomness to generate the schedule and as such it may not always give the best solution or be able to resolve every day. If you find the schedule isn't particularly balanced on the first try, run it a few more times and select the best output. You can also manually edit the text file before committing to Google Calendar.

Committing a Schedule to Google Calendar

To commit a schedule to Google Calendar you need:

1. A Google account
2. Edit access to the google calendar you want to update / commit to.
3. The Calendar ID for the google calendar you want to update / commit to. You can find this in the Calendar Settings.
4. A schedule file in the exact format as those generated by this program (see above). That is the file should be of the form:

<day of week> : <name> : <date in YYYY-MM-DD format>

For example

thursday : John : 2018-02-15
friday : Jill : 2018-2-16

5. A Google Calendar API Key. (Eventually I'll try to package this so you don't have to do that, but as of now this is the best I can do.) Follow the steps here: https://developers.google.com/google-apps/calendar/quickstart/python to get your own Google Calendar API key. Make sure you move the client.json file to the same folder as scheduler.py.

Once you've generated a schedule (see above) you can commit it to google calendar using:

	$ python scheduler.py --commit --infile schedule_file.txt

Or

	$ python scheduler.py -c -i schedule_file.txt

The commands above will eventually prompt you to enter a Google Calendar ID. You can also enter the Google Calendar ID as a command line flag

	$ python scheduler.py -c -i schedule_file.txt -cal <CALENDAR ID>

Where <CALENDAR ID> is the calendar ID copied from the Calendar Settings.

Analyzing a Schedule

If you've manually created or changed the schedule you may want to verify you still meet the RAs' preferences and that the number of weekday / weekend duties are balanced. The analyze.py program has a few features that will help with this.

$ python analyze.py -cr -pf preferences_file.txt -sf schedule.txt

Checks that the schedule in schedule.txt conforms to the restrictions found in preferences_file.txt. If not, it will print the dates of any errors.

$ python analyze.py -gs -sf schedule.txt

Prints a summary of weekday / weekend duties for each RA based on the schedule in schedule.txt.

$ python analyze.py -cr -gs -pf preferences_file.txt -sf schedule.txt

Does both of the tasks above.

Issues or Updates

If you spot some kind of bug you can approach me directly or open an issue on github.

If you'd like to make a contribution to this repo please submit a pull request.

To Do

• Look at two-back conflicts.
• Abstract for more general schedules.
• Develop "best of batch" scheduling
• Limited forward checking / edge consistency
• Heuristic for resolving conflicts