Fin-dashboard - Your financial markets dashboard

Fin-dashboard is a web application that provides a dashboard for financial markets. It is built using Streamlit. It leverages the facebook's Prophet ML library to help predict the stock/index prices for next n days/months/years.

To access the web application visit Fin-dashboard.

Fin-dashboard features

Fin-dashboard provides a dashboard with the following features:

1. Automatic detection of the user’s country to show predictions for the top five, country-specific stocks

2. Ability for users to select long- or short-term predictions and to specify the number of forecasted days

3. Ability for users to choose other stocks and other stock exchanges

4. For long term predictions, a comparison of the model’s predicted price with the analyst’s target price

5. For short term predictions, a sentiment analysis of the stock news to enable visualization of its impact on stock price

6. Correlation of a stock with the index

7. Long term predictions for stock exchanges of different countries

To read a more detailed guide on how to use and deploy the application, refer to Edge Analytics with SUSE Rancher: Finance - Market Predictions.

To see the presentation and explanation of the Fin-dashboard project, refer to this video.

File Structure

Fin-Dashboard is composed of the following Python modules:

1. geo_location.py: Determines the country where the user is located to give country specific information.

2. index_correlation.py: Correlates the stock with its parent index.

3. main.py: Contains all the Streamlit configurations/page setup and uses the other files to create the application.

4. plot_charts.py: Contains the Prophet ML model for predicting stock prices. Contains two python functions - one for long term and other for short term predictions.

5. sentiment.py: Collects stock news from Google News and uses the VADER sentiment analysis tool to determine the stock’s trend.

6. stock_recomendation.py: Extracts the analyst’s price target of a stock from Yahoo Finance.

Steps to deploy the fin_dashboard app locally

1. Install necessary libraries using pip install -r requirements.txt
2. Run streamlit run main.py and access the application on localhost:8501

Steps to deploy the fin_dashboard as a container

1. Pull the image from docker hub sudo docker pull navin772/fin_dashboard:latest or build the image yourself from the Dockerfile using docker build .
2. Start a new container using the image sudo docker run -p 7000:8501 <image_id>
3. Access the streamlit app on localhost:7000

Steps for deploying on k8s cluster

1. Run kubectl apply -k yaml_files and kustomize will apply all the yaml files to the cluster currently configured with kubectl.
2. Access the app on <NodeIP>:<NodePort>