vprof

vprof is a comprehensive Python profiler that provides visual analysis of program execution, helping developers identify performance bottlenecks, memory usage patterns, and code hotspots. Developed by Vlad Degtev, vprof integrates four profiling modes into a single tool: CPU profiler (measuring execution time of each function with call graph visualization, showing inclusive and exclusive time), memory profiler (tracking memory allocations and deallocations over time, identifying memory leaks and peak usage points), line profiler (measuring time spent on each individual line of code, essential for micro-optimization), and code heat map (colorizing source code lines by execution frequency and time spent). The tool generates interactive HTML reports that can be viewed in any web browser, featuring call graphs (with d3.js visualization showing function relationships and call frequency), flame charts, and sortable tables. vprof can profile both scripts run from the command line and imported modules, supporting CPython 3.8+. It uses cProfile and Yappi as backends for CPU profiling, and memory_profiler tracemalloc for memory analysis. The tool is particularly useful for analyzing long-running applications, data processing pipelines, and web server performance. Licensed under the BSD 2-Clause License.