使用Visual Studio和大型数据集进行软件测试(时间序列) [英] Software Testing with Visual Studio AND Large Datasets (Time Series)

问题描述

我正在寻找有关使用大型数据集(使用Visual Studio中的本机单元测试工具)在Visual Studio中进行测试的很好的参考(参考网页或书籍). (我有Visual Studio 2008 Professional.)

互联网上有一些关于使用数据库作为测试输入的文章.在这种情况下，仅数据库用于输入(单个)参数值.我的项目虽然涉及一些方法，但这些方法需要对较长时间序列(共同基金收益)进行计算，而多个共同基金至少需要36个月的收益.我正在寻找的是有关如何测试该项目的教程(使用一个或几个固定数据集).

I''m looking for a good reference (to a webpage or book) on testing in Visual Studio with large datasets (with the native unit testing tools in Visual Studio). (I''ve got Visual Studio 2008 Professional.)

There are some articles on the internet on using a database as input for testing. Only the database in that case is used for input of (single) parameter values. My project though involves methods that do calculations on long time series (of mutual fund returns) that need at least 36 months of returns for multiple mutual funds. What I''m looking for is a tutorial on how to test this project (using one or a couple of fixed data sets).

Any references?

推荐答案

测试数据集有多大?您是否必须测试数据库中每一个数据?通常，您应该能够根据特定场景创建一些测试用例，而不必进行完整的回归测试.

函数签名是什么样的?您是否能够创建包装函数，将数据从db插入处理该数据的函数?如果这样做，是否可以对包装函数进行单元测试?我认为数据库/大型数据集方面不必由您的测试来管理，而是一个功能?

How large is the testing dataset? Must you test every single peice of data in the DB? Typically you should be able to create some test cases based on certain scenarios and not have to do a full regression test.

What does the function signature look like? Are you able to create a wrapper function that plugs the data from the db to the function that processes it? If you do this, could you just unit test the wrapper function? I''m thinking the database / large dataset aspect does not have to be managed by your test, but rather a function?

这篇关于使用Visual Studio和大型数据集进行软件测试(时间序列)的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！