Method for testing statistical function and terminal

Abstract

The invention relates to the field of testing, in particular to a method for testing a statistical function and a terminal. The method comprises the steps: obtaining total data from a production environment database corresponding to a to-be-tested application program; wherein the data in the production environment database is generated by real user operation; testing algorithm logic correctness ofeach statistical function in the to-be-tested application program by a preset test script according to the total data to obtain an algorithm logic test result; testing the business logic correctnessamong the statistical functions in the to-be-tested application program according to the total data to obtain a business logic test result; and generating a test report according to the algorithm logic test result and the service logic test result. The method for testing a statistical function can improve the accuracy of the test statistical function.

Description

Technical field [0001] The invention relates to the field of testing, in particular to a method and terminal for testing statistical functions. Background technique [0002] With the update and iteration of Internet technology, more and more companies rely on big data statistical analysis platforms to analyze their own product operation and maintenance data and make operational decisions based on this. Because of this, the accuracy of the data in the statistical analysis platform is very important. [0003] Big data statistical analysis platforms generally have the following characteristics: 1. There are many indicator dimensions and the statistical logic is independent of each other. For example, the number of new users and active users can be counted from the overall dimension, and the two indicators can also be counted from the regional dimension. These two indicators can be counted from the channel dimension; 2. There are various types of access products, and there are many ty...

AI Technical Summary

Problems solved by technology

However, there are many access products, and the basic data types of different products are still different. Therefore, in the test traversal process, it takes a lot of time and manpower to complete all product data verification. The amount of data is very large. The single basic data volume of single-dimensional statistical indicators can reach ten thousand levels;

[0006] (2) Manual testing cannot traverse all the data of the entire application in a short period of time, so there is a risk in the accuracy of product data that has not been traversed;

[0007] (3) Manually test multiple applications, adopt the same test method, there are more repetitive tasks, and the execution ability is doubled;

Therefore, in the testing process, in addition to verifying the accuracy of single-dimensional indicators, it is also necessary to compare the data accuracy of the same type of indicators in different dimensions. However, due to the limited energy of manual testing, it is impossible to complete the data comparison of all applications in a short time.

Images