Skip to content

Latest commit

 

History

History
59 lines (37 loc) · 2.92 KB

README.md

File metadata and controls

59 lines (37 loc) · 2.92 KB
Raw

TOST - Trip Outlier Scoring Tool

Backend

Requirements

We are using p3 venv

All the dependencies are saved as an environment file on the root directory. The file name is environment.yml. You can create the environment with all the dependencies using conda.

conda env create -f environment.yml

This will create a conda environment named tost. You can then verify the new environment is installed correctly running conda env list

To activate the environment run conda activate tost

Dataset and Database

Requirements

  • Postgres - PostgreSQL, A relational DB solution
  • Postgis - A Postgres extention to work with geometries and geographical datatype

The dataset used in this work is composed of trips between the ports of Houston and New Orleans between 2009 and 2014. This dataset is composed of two csv files.

The sample data is available at this google drive shared folder

A PostgreSQL (aka Postgres) database is used to store the raw trip and vessel data for ease of pre-processing and querying. To populate the database from the raw AIS data, the below steps need to be followed -

  1. Create a database in Postgre named tost-db
  2. Run the sql create statements in the /database-scripts directory in the given sequence in the filename prefix (v01-v06). Running all these create statements will create all the required table in the tost-db database
  3. When the tables are there, to fill in the tables from the raw AIS data in the available csv files, first download the csv files from the shared Google drive folder. Then, in the /scripts directory there are python scripts to read in the csv files and write in the database table.
    Change the username and password to your Postgres username and password in the db_connection.py file.
  4. In script1.py and script3.py set the csv_file_path variable to the directory where you stored the raw trip and vessel data csv files.
  5. Running all the python scripts in the sequence script1 to script9, will enumerate on the raw data and populate the table. These scripts will also interpolate the trip trajectory's missing data, divide the trip into different segments, calculate trip specific values for the segments and store them in the trip v2 table.

Server

For the sake of consistency we are using the same language for the server as for the scripts. The server-side is a python flask application.

Requirements

Flask

pip install flask

TODO

  • Add different segment sizes
  • Add itinarary table which will have a origin and destination
  • Adding an interpolation filter for the Tablelense view.