python - 使用 Enthought 和安装单个库进行比较和对比

Question

使用 Enthought 和安装单个库进行比较和对比。我不清楚每个选项提供的好处。

Answer 1

Enthought allows for a lot of scientific and mathematic libraries to be installed through the ease of a UI. This however, means you could be installing a good deal of libraries that are irrelevant to your needs. It does however come with a suite for visualizing your code.

Although it takes more time, going out and looking at different releases of specific libraries that are more centric to your requirements would probably be more beneficial. Mainly NumPy, SciPy and PyTables would be large ones to go and install yourself.

In the end, it's all up to you if you want to download bulk with a lot of good stuff easily... with some unnecessary baggage; or, you could research and install specific libraries to fit your needs so you know exactly what to use, and how... for 'cheaper' on a data level anyways.

Personally, I'd probably install Enthought.

On an unrelated note, this question isn't horribly fit for this site. --> https://stackoverflow.com/faq#dontask Mentions that "What's your favorite_____.... type questions are not very good.

Answer 2

这完全取决于您希望使用哪些 python 库。我自己在 OS X 上的经验是，大多数 python 模块都需要编译工作，但我偶尔会遇到问题。特别是如果我需要匹配编译为通用二进制的系统 Python 的体系结构。老实说，这些天我只是在使用 Enthought，特别是因为我是一名学者并且许可证是免费的——当我很好地询问他们时，他们甚至可以让我访问 64 位安装程序。

当一位老板决定将我的代码移动到某些 Windows 机器上运行时，我也对我的决定感到高兴……我很高兴我可以将它们指向单个 Enthought 安装程序，而不必通过安装 python 牵着他们的手+ 外部库。

总而言之，Enthought 可能对您免费，也可能不免费……如果它是免费的，那么即使您在具有良好软件包存储库的 Linux 发行版上进行开发，也几乎肯定值得使用。如果您必须为 Enthought 付费，那么您将不得不评估它是否物有所值，只需记住将您自己的时间成本考虑在内。

如果您引用您想要使用的特定软件包和您的平台，我可以更好地猜测 Enthought 可能对您有多大的优势。