Web app to download, analyze and save currency exchange rates for specified currency pairs.

App utilizes https://api.nbp.pl/ to get currency exchange rates.
Data is stored in all_currency_data.csv file.
Server side is made with Python & Flask. Prepared endpoints work with that .csv file to provide data for the frontend made with React.
User can select currency pairs and download them as .csv or save it to the server.

1. Git clone repo

git clone https://github.com/saworz/recruitment-task.git

2. Move to repo dir

cd recruitment-task

3. *(To run with docker) Run docker compose

docker compose up --build

4. *(To run via terminals) Or run server and front separately

cd backend
pip install -r requirements.txt
python app.py

cd frontend
npm install
npm start

Task requires both data manipulation and user interaction, so I've decided to create web app that runs with Flask server and React frontend. Server is pretty compact and contains only few endpoints and Django would be overkill for such a simple task so I've chosen Flask. I could stick with pure Python and create visualization with Streamlit, but I've started learning MERN stack so I decided to show of some of my frontend skills with HTML, CSS, JS, React and Typescript.
You have also mentioned Docker in job requirements so I have packed app in docker network.

I have created NbpFetcher class that handles fetching data from nbp api. Requests are made for specified list of currencies named with Currency Codes ISO 4217 convention.
Fetched data is converted to pandas dataframe and then saved to .csv file. It utilizes pandas because it makes handling column data and .csv converting easy.

User can select currency pairs he wishes to get exchange rates for by clicking checkbox on the right side of the currency pairs. Each currency pair div is created by mapping list of available currencies fetched from the /api/get_currency_types/ endpoint.

Dataframe mentioned in paragraph 2 saves the data to .csv file named "all_currency_data.csv" file. By selecting data with checkboxes user can choose currency pairs exchange rates to save, either locally or to the server. Download as .csv button downloads .csv file with selected pairs and Save to server saves the same file to the server.

Success messages, errors and info are displayed for the user with react-toastify library. It makes creating popup messages pretty simple and it's easy to customize them.

Errors occurring while fetching data or communicating with backend are displayed with react-toastify. They appear in the upper-right corner and provide simple error description. More precise error messages are displayed with Python's logging module and in browser's console. Data types are validated using Pydantic and Flask-Pydantic for api requests and responses

/api/analyze_data/ endpoint analyzes data for provided list of currency pairs. It also utilizes pandas library to find average, median, min and max values easily. They are returned to the frontend and displayed as info popups in the upper-left side of the screen. Results are visible on the image in paragraph 5.

APScheduler takes care of fetching nbp exchange rates data once a day at 00:00. fetch_nbp_api function is used as a cyclic job and the scheduler is started by running app.py script.