I want to create a real time map dashboard using JS + python + flask + amcharts, most of the code is already done, I just need the final part which is connecting amcharts with new locations in real time that comes from an API.
Description of the overall system:
I'm creating a fraud detection solution. The code can be found in the branch fraud_map of this repo: [login to view URL] The first part is a machine learning model that learns when a credit card transaction is fraudulent or not. I created an API in flask to access the machine learning model. This part is finished.
The second part is the visualization in amcharts of a world map of showing the locations of fraudulent and fair transactions. When I make a call to the API, with information of the credit card transaction, the model predicts if the transaction is fair or fraudulent. This part is done, as it is shown in this code: [login to view URL] The next step is to show in the map the location of the fraudulent transaction. This is the part that I need help with. If the transaction is fraudulent, I need a new point in the map of color red. If the transaction is fair, the new point will be blue.
Description of the parts finished:
1) machine learning model that predicts whether a credit card transaction is fair or fraudulent.
2) API that given an input with a card transaction, calls the machine learning model and returns a prediction in json.
3) Map using amcharts of the world showing the coordinates of cities. The points are in red and use a pulse to highlight the location.
Description of the work to do:
1) Add the needed code to allow a real time update of the map. When I do a new call to the API, calling to http://localhost:5000/predict_map with a json of the creadit card information, I want to create a new point on the map representing the location where the transaction is being done. If the transaction is fair, the point has to be blue, if it is fraudulent, the point has to be red. Ideally I would like to do it with a websocket.