https://github.com/handshake/pystradamus

Evidence-based scheduling tool for Jira
https://github.com/handshake/pystradamus

Last synced: 24 days ago
JSON representation

Evidence-based scheduling tool for Jira

• Host: GitHub
• URL: https://github.com/handshake/pystradamus
• Owner: handshake
• License: mit
• Created: 2014-03-28T17:26:19.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2014-04-11T18:45:18.000Z (over 10 years ago)
• Last Synced: 2024-08-04T04:04:08.973Z (4 months ago)
• Language: Python
• Size: 203 KB
• Stars: 13
• Watchers: 6
• Forks: 2
• Open Issues: 1
• Metadata Files:
  • Readme: README.rst
  • License: LICENSE.txt

Awesome Lists containing this project

• starred-awesome - pystradamus - Evidence-based scheduling tool for Jira (Python)

README

        
Pystradamus

===========

Pystradamus is an evidence-based scheduling tool that mines Jira tickets to build probability curves for future work.

Installation

------------

To install via pip (the recommended method)...

.. code-block:: bash

    $ pip install git+https://github.com/Handshake/pystradamus

Now have pystradamus dump a generic config file into your home directory

.. code-block:: bash

    $ pystradamus config > ~/.pystradamus.cfg

Next, edit your specific Jira settings in the config we just

made. Pystradmus will use this configuration from then on. You can also

override the config to use at run time with the `-c` flag.

Usage

-----

Pystradamus works in two stages: Refresh and Predict. The refresh stage

pull historical data from your Jira instance for a particular usage. Example...

.. code-block:: bash

    $ pystradamus history -r joe

This will find all tickets that had an estimate in the past and build

a local sqlite3 database of these estimates as well the time spent

"in-progress" per ticket in Jira.

Predict mode is what projects the currently open and assigned tickets

for this user into the future. For example after having refreshed user

`joe` in the previous step we can now predict his currently open and

assigned workload like so...

.. code-block:: bash

    $ pystradamus history -p joe

    HS-3503 [0.05] Update storefront promo edit preview to match actual dash

        50% chance: 2014-04-11 19:19:59.466628

        95% chance: 2014-04-11 21:24:51.305642

    HS-3323 [0.1] Custom help page for buyers

        50% chance: 2014-04-13 01:10:35.751628

        95% chance: 2014-04-18 19:13:52.397642

    HS-3411 [0.1] Buyer "My Settings Page"

        50% chance: 2014-04-14 02:24:05.444628

        95% chance: 2014-04-25 17:02:53.489642

    HS-3295 [0.1] Storefront Order Screen Rework

        50% chance: 2014-04-15 03:37:35.137628

        95% chance: 2014-05-02 14:51:54.581642

    HS-3310 [0.05] Storefront explanation for multiple ship dates

        50% chance: 2014-04-15 04:15:35.758628

        95% chance: 2014-05-03 11:24:50.788642

What you've just been given is Joe's next 5 tickets, with their 50%

and 95% confidence fits to the calendar. The more history for a user

the better the predictive results.