Parking Data

Analysis, cleaning, and visualization of data used for and produced by research leveraging AI/ML to predict parking availability

Parking Availability Forecasting Model - https://ieeexplore.ieee.org/document/9071688

---

GETTING IT RUNNING:

-You must install node.js 'http-server' module and add it to your path.

+npm install http-server

-Run localserver.bat and open 'localhost:8080' in browser

-Otherwise, find another way to host a local server

---

GUI:

Buttons are self-explanatory

Push 'Enter' after altering text box for time selection to update visualization

Slider located on top left for changing transparency (Updated visualization on mouse-click-up)

---

NOTES/WARNINGS:

WILL NOT LOAD WITHOUT BEING RUN ON A SERVER

---

What are the files in '/KansasCityData/'?

Go to 'comparisons' folder in the Google Drive for a visual representation

Flow:

    sensity_events.csv  --------->  legibleParking.csv  --------->  cleanedParking.csv & formatted_cleanedParking.csv --------------------> sensorInformation.csv

                        legible.py           |          clean.py                                                       findSensorBounds.py

                                             |               ^

                                collisions.py|               |

                                             V               |

                                        overlapmatrix.txt ----

legible.py: Cleans the data's format (just string manipulation)

collisions.py: produces a matrix of collision events > 20% overlap

clean.py: uses the overlaps to produce a new version of the data with X% removal bias (currently 0, or 100% removal of overlaps > 20%). Also creates a version in the same format as the original sensity_events.csv data