Abstract :
With a wide range of Python language in developing programs, More and more programmers choose to use the Python language for systems development, it gradually becomes scientific computing, web and games´ Choice Awards. However, the performance of python is always a headache for developers. For reasonable selection of functions in base library, the usage of third-party plug-ins´ functions and methods, and the design of custom functions, the problem whether they are the best choices for general developers is difficult to make a positive answer. After the system´s performance bottleneck occurs, it is particularly important to determine where to tune and how to tune. Through the analysis and dynamic tracking of source code, with the built-in method in Python, we can get information about the system to be optimized, included: functions, grammatical structures, running time of each function, the relationship between function calls etc. This information provides an effective basis for further optimization of the system. Experimental results show that the system optimized by the tool has a significantly improvement.
