Skip to content

Latest commit

 

History

History
175 lines (146 loc) · 4.59 KB

README.md

File metadata and controls

175 lines (146 loc) · 4.59 KB

Andromeda Backend Webserver

This server allows users to upload CSV files and perform dimensional reduction operations on the data.

Requirements

  • Python 3.8+

Setup

Install depenencies within a python virtual environment:

pip install -r requirements.txt

Running

By default flask performs CORS checking which is useful in production. When developing locally this can cause problems when using the API from a web browser. To disable CORS checking functionality run the following:

export ANDROMEDA_DEV_MODE=Y

Within your python virtual environment run:

flask --app main run --debug

Run Tests

Within your python virtual environment run:

python -m unittest

Run Integration Test

You will need jq installed to run this test. The flask app needs to be running in another terminal.

Run the following command:

./itest.sh

API

Definitions

  • dataset_id - Unique identifier for a CSV file uploaded to the API
  • column_name - name of a column in the CSV file
  • label unique name of a particular image
  • x is a number between -1.0 and 1.0 representing the x location of the image
  • y is a number between -1.0 and 1.0 representing the y location of the image
  • url URL of an image

Create a dataset

Upload a CSV file and return an id for use with dimensional-reduction and inverse-dimensional-reduction endpoints.

  • POST /api/dataset/
    • Input
      • file: CSV file to create a new datset for
    • Output 
      { "id": "<dataset_id>" }

Download a dataset

  • GET /api/dataset/<dataset_id>?filename=
    • Input
      • dataset_id: id of the dataset to download
      • filename: optional filename to use for Content-Disposition
    • Output
      • dataset CSV content

Perform dimensional reduction

Calculate image coordinates for the given weights.

  • POST /api/dataset/<dataset_id>/dimensional-reduction
    • Input 
      { "weights": {"<column_name>": <weight_value>, ...} }
    • Output 
      { 
    "id": "<dataset_id>",
    "weights": {"<column_name>": <weight_value>, ...},
    "images": [{"label": "<label>", "url": "<url>", "x": "<x>", "y": "<y>"}, ...]
}

Perform inverse dimensional reduction

Calculate weights for the given image coordinates.

  • POST /api/dataset/<dataset_id>/inverse-dimensional-reduction

    • Input

      { "images": [{ "label": "<label>", "x": <x> "y": <y> }, ...] }

    • Output

      {
    "id": "<dataset_id>",
    "weights": {"<column_name>": <weight_value>, ...},
    "images": [{"label": "<label>", "url": "<url>", "x": "<x>", "y": "<y>"}, ...]
}

Fetch iNaturalist observations as JSON data

Read all observations for iNaturalist user and return as JSON.

  • GET /api/inaturalist/<inat_user>
    • Output 
      {
  "data": [
    {
      "Date": "...",
      "Image_Label": "...",
      "Image_Link": "...",
      "Lat": 42.9349293461,
      "Long": -88.0380119262,
      "Place": "...",
      "Seconds": 57785,
      "Species": "White-tailed Deer",
      "Time": "...",
      "User": "..."
    },
    ...
  "user_id": "...",
  "warnings": [
    "missing_lat_long"
  ]
}

Fetch iNaturalist observations as CSV data

Read all observations for iNaturalist user and return as a CSV file.

  • GET /api/inaturalist/<inat_user>?format=csv
    • Output 
      Image_Label,Image_Link,Species,User,Date,Time,Seconds,Place,Lat,Long
p1,https://static.inaturalist.org/photos/12647253/medium.jpg,White-tailed Deer,lhouse,2017-12-31,16:03:05,57785,"Whitnall Park, Hales Corners, WI, US",42.9349293461,-88.0380119262
...

NOTE: Since the result is a CSV file no warnings will be returned.

Create a dataset for some iNaturalist observations

Creates a dataset for iNaturalist observations returns JSON with the dataset id and a URL for downloading

  • POST __/api/inaturalist/<inat_user>/dataset?add_sat_rgb_data=<add_sat_rgb_data>&add_landcover_data=<add_landcover_data>
    • Input
      • inat_user: iNaturalist username to create a CSV for
      • add_sat_rgb_data: Pass 'true' to include RGB data in resulting CSV
      • add_landcover_data: Pass 'true' to include landcover data in resulting CSV
    • Example Output

{ "id": "...", "url": ".../api/dataset/...", "warnings": [...] } ```

Get column configuration settings

Retrieve configuration about ancillary column names

  • GET /api/column-config/
    • Output 
       "ancillary_columns": [
    "sat_Lat-Center", ...
 ]