How can you determine the change in program performance after making modifications?

The following approaches are used to determine performance changes:

• Profiling: Using tools such as Android Profiler in Android Studio to monitor CPU, memory, network, and energy consumption. This helps identify bottlenecks in the code.

  // Example of simple timing measurement of a code block
  val startTime = System.nanoTime()
  // Code whose performance needs to be measured
  val endTime = System.nanoTime()
  val duration = endTime - startTime // Difference in nanoseconds
  

• Benchmarking: Creating test scenarios that measure the execution time of specific operations or code sections in a controlled environment.

• Performance metrics: Tracking key indicators on real users or test devices:

  • App startup time
  • Frames per second (FPS)
  • Memory usage
  • Battery consumption
  • Network usage

• Automated testing: Writing tests that include performance checks and running them on CI/CD pipelines. This allows tracking changes over time.

• A/B testing: Comparing the performance of two versions of a program on different user groups.

• Real User Monitoring (RUM) tools: Using services like Firebase Performance Monitoring, New Relic, Sentry to collect performance data on users' devices.

By comparing these data before and after changes, one can draw conclusions about the impact of the modifications on performance.