Statistics about actual program performance are always your best first tool in the pursuit of making applications faster. The good news is, Python ... profiling apps in production with live data ...

Tools included with the stock installation of the Python interpreter can provide you with detailed feedback about which parts of your program are slow, and offer some hints about how to speed them up.

Add interactive elements to your plots, such as hover-over information, zoom features, and clickable legends, which can be achieved through Plotly or Shiny apps in R. Use profiling tools like py-spy ...