|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/streamlit/
课程评论：没有评论
课程名称：完整的数据科学应用指南（使用Streamlit） 课程概述：分析数据和构建机器学习模型是一方面，而将这些分析和模型打包成可分享的应用则是另一项挑战。本课程旨在教授您使用Streamlit构建和分享数据应用的最快、最简单的方法。您无需具备构建前端应用的经验。课程将覆盖以下内容： - Python速成课程 - NumPy速成课程 - Streamlit简介 - 在Streamlit中整合Matplotlib和Seaborn - 在Streamlit中使用Altair和Vega-Lite - 理解所有Streamlit小组件 - 上传和处理文件 - 构建图像处理应用 - 开发自然语言处理应用 - 在Streamlit中整合地图 - 实现Plotly图表 - 认证您的应用 - 在Streamlit中布局您的应用 - 使用Streamlit组件进行开发 - 部署数据应用 为什么选择Streamlit：尽管还有其他库可以用来构建数据应用，但Streamlit有其独特优势，无需前端经验，所有代码都可用您熟悉的Python编写，便于与小组件如滑块实现交互，并且快速易于部署，兼容大多数数据科学框架。 Python脚本：在Streamlit中构建数据应用时，您不会离开Python编辑器，因为所有功能都在Python中编写。与其他框架不同，Streamlit避免了必须编写HTML、CSS和JavaScript代码的麻烦。 互动性：添加互动功能非常简单，Streamlit提供的小组件可以轻松实现用户交互。 部署：分享Streamlit应用也很轻松，您可以轻松将应用部署到Heroku和AWS，或者使用Streamlit Sharing一键部署，只需请求访问即可。 兼容性：Streamlit与流行的数据科学库兼容，您可以使用Matplotlib、Seaborn、Altair、Plotly和Bokeh等工具进行可视化。此外，Pandas和NumPy支持数据清洗与处理。 Streamlit组件：若需要的功能Streamlit没有，可以查看Streamlit组件页面，利用社区构建的第三方组件，或自己编写共享。 课程结束时，您将建立多个可纳入数据科学作品集的应用，获得一项新技能，且课程附带30天无条件退款保证。立即报名，如果不满意将全额退款。
Analyzing data and building machine learning models is one thing. Packaging these analyses and models such that they are sharable is a different ball game altogether. This course aims at teaching you the fastest and easiest way to build and share data applications using Streamlit. You don't need any experience in building front-end applications for this. Here are some of the things you can expect to cover in this course: Python Crash CourseNumPy Crash Course Introduction to StreamlitIntegrating Matplotlit and Seaborn in StreamlitUsing Altair and Vega-Lite in Streamlit Understand all Streamlit WidgetsUpload and Process FilesBuild an Image Processing ApplicationDevelop a Natural Language Processing Application Integrate Maps with Streamlit Implement Plotly Graphs Authenticate Your ApplicationsLaying Out your Application in StreamlitDeveloping with Streamlit ComponentsDeploying Data Applications Why StreamlitThere are several other libraries that can be used for building data applications. That said, why should you consider Streamlit:No front-end experienced requiredWrite everything in what you already know - PythonEasy to weave in interaction with widgets such as slidersQuick and easy to deployCompatible with most data science frameworksNo front-end experienced requiredIf you were to build a data app with Flask and or Django, then knowledge in front-end tools such as HTML & CSS as well as Javascript is a must. However, in Streamlit, all this is done using Streamlit widgets. For example, a drop-down can easily be achieved using the selectbox widget. Other HTML tags such as input boxes and buttons are also achieved using simple Streamlit widgets.Python ScriptingWhen building data applications in Streamlit, you never leave your Python editor. This is because is scripted in Python. It is, therefore, very advantageous since you keep working in a language that you are already familiar with. If this was done in other Python frameworks, then writing HTML, CSS, and Javascript code would be unavoidable.InteractivityAdding interaction to Streamlit applications is very simple. Streamlit provides widgets that one can use to weave interactivity to your application. For example, one can use the date input widget to filter their data. Select boxes and sliders can also be used to achieve the same.DeploymentSharing Streamlit applications is very easy. One can easily deploy to the likes of Heroku and AWS. However, one can also deploy their app on Streamlit Sharing by the click of just two buttons. All you have to do is to request access. Your Github email address will then be linked to Streamlit Sharing. Once this is done, you can deploy any Streamlit project available on your Github account.CompatibilityStreamlit is compatible with the most popular data science libraries. For example, you can perform visualizations in Streamlit with the tools that you are already used to. The visualizations libraries supported include:MatplotlibSeabornAltairPlotlyBokehYou definitely need to perform data cleaning and wrangling before visualizing your results. Pandas and NumPy are supported so that you can achieve this.When it comes to machine learning, you can deploy models built with the popular libraries that you are already used to. This is because Keras, TensorFlow, and PyTorch are supported out-of-the-box.Streamlit ComponentsIn the event that you need a functionality that is not supported by Streamlit the first place to look is the Streanmlit Components page. Streamlit Components are third-party functionalities that have been built by the community. The components can be installed via pip and used immediately in your project.Streamlit ComponentsThe beauty of it is that you can also write your own components and share them with the community.At the end of the course, you will have built several applications that you can include in your data science portfolio. You will also have a new skill to add to your resume. The course also comes with a 30-day money-back guarantee. Enroll now and if you don't like it you will get your money back no questions asked.
