What are the classic data visualization libraries in Python?

Data visualization is an effective means to display and understand data, and the commonly used Python data visualization libraries are as follows:

The first python visualization library, which many other libraries are based on or directly called, can easily get an overview of the data, which is very powerful, but also very complex.

The biggest difference between matplotlib and matplotlib is that the default drawing style and color matching have a modern aesthetic.

GGPLOT2, which is based on R's library, also makes use of the concept derived from the Image Grammar, which allows different layers to be superimposed to complete a picture, and is not suitable for making very personalized images, sacrificing the complexity of the image for the simplicity of operation.

It's very similar to ggplot, but differs from ggplot in that it's entirely python-based and not referenced from r. The advantage is that it can be used to create interactive, network-ready diagrams.

Charts can be output as JSON objects, HTML documents, or interactive web applications.

It can be used through a Python notebook and is dedicated to the creation of interactive diagrams like Bokeh, but provides several chart types that are almost impossible to find in other libraries, such as contour plots, tree diagrams, and 3D diagrams.

As with bokeh and plotly, interactive images are provided that can be embedded directly into a web browser. The main difference with the other two is that the charts can be exported to SVG format, all the charts are encapsulated into methods, and the default style is also beautiful, and it is easy to make beautiful charts with a few lines.

A toolbox for making maps and geographically relevant data. It can be used to create a variety of maps, such as equivalent area maps, heat maps, point density maps, etc., and pyglet must be installed before use.

Quickly assess missing data in the form of images, sort or filter data based on completeness, or correct data based on heat maps or treemaps.

1、matplotlib

MattPlotlib is the originator of many data visualization libraries in Python, and its design style is very close to the commercial programming language MATLAB designed in the 80s of the 20th century, with many powerful and complex visualization functions. MattPlotlib includes several types of APIs that can plot and customize graphs in a variety of ways.

2、seaborn

Seaborn is an advanced encapsulated visualization library based on MattPlotlib, which supports an interactive interface that makes it easier to draw icons and more attractive colors for drawing rich and diverse statistical charts.

3、ggplot

ggplot is a library based on matplotlib and designed to make matplotlib visualization more appealing in a simple way, drawing graphs in the form of overlay layers. For example, the layer where the axes are drawn, then the layer where the points are located, and finally the layer where the lines are located, but it is not suitable for personalized graphics. In addition, GGPocta2 prepares an interface for the R language, in which the API is not suitable for Python, but is suitable for the R language and is very powerful.

4、bokeh

Bokeh is an interactive visualization library that can be presented using a web browser and can be used to quickly and easily transform large data sets into high-performance, interactive, and simple-structure charts.

5、pygal

PyGal is a library of scalable vector charts for generating SVG-formatted charts that can be opened in the browser and automatically zoom on screens at different scales for easy user interaction.

6、pyecharts

pyecharts is a library for generating echarts, and the generated echarts have been recognized by many developers for their good interactivity and exquisite design.

1.MattPlotlib: It is the originator of many data visualization libraries in Python, and its design style is very similar to the commercial programming language MATLAB in the 80s of the 20th century, with many powerful and complex visualization functions; There are also several types of APIs that allow you to draw and customize icons in a variety of ways.

2.Seaborn: is an advanced encapsulated visualization library based on Mattplotlib that supports an interactive interface that makes it easy to draw charts and make the colors of the charts more attractive.

3.ggplot: is a library based on matplotlib and designed to improve the visual appeal of matplotlib in a simple way, using the form of overlay layers to draw graphs, such as first draw the layer where the coordinate axis is located, then draw the layer where the points are located, and finally draw the layer where the lines are located, but it is not suitable for personalized graphics.

4.Boken: is an interactive visualization library that can be presented using a web browser, and can be used to quickly and easily transform large data sets into high-performance, interactive, and simple-structure charts.

5.pygal: is a scalable vector icon library for generating browser-ready SVG-formatted charts that automatically scale to a screen at different scales for easy user interaction.

6.pyecharts is a library for generating echarts, and the generated echarts have been recognized by many developers for their good interactivity and exquisite design.

This article is the second summary tutorial of the "Data Frog Three-month Intensive Course", if you want to knowData Frog Community, which can be readStudy advice for the DataFrog communityFor us data analysts, we must not only understand the meaning behind the data, but also show the meaning of the data to the boss more intuitively. So, let's learn this indispensable skill together.

Before drawing, we import the package and generate the dataset.

Let's start with the dataset used.

Line charts are the most common graphs we use to observe the trend of the calendar, and you can see the trend of the data with a certain variable, and the parameter is kind= by default"line"The type of representation chart is a line chart.

For discrete data such as categorical data, where you need to see how the data is distributed across categories, you can use a bar chart. We draw a column for each category. At this point, you can set the parameter kind to bar.

A bar chart is a graph obtained by flipping a vertical bar chart 90 degrees. Like bar charts, bar charts can have one or more sets of data.

Horizontal bar charts are very handy when category names are long, because the text is written from left to right, in line with the reading order of most users, which makes our graphics easy to read. And the histogram is not well displayed when the category name is very long.

A histogram is a special form of a column chart, and we choose a histogram when we want to see the distribution of a data set. The variables of the histogram are divided into different ranges, and then counted in different ranges. In the histogram, continuous, continuous bars between bars imply numerical continuity.

A box plot is used to display the descriptive statistics of a dataset, that is, the [quartile], and the upper and lower ends of the line represent the maximum and minimum values of a certain set of data. The upper and lower ends of the box represent the top 25% and 75% of the values in the data. The horizontal line in the middle of the box indicates the median.

You can now set the parameter kind to box.

If you want to draw a scatter plot, you can set the parameter kind to scatter and specify x and y. Scatter plots allow you to explore the relationships between variables.

A pie chart is a percentage of a set of data in terms of area, and the parameter kind can be set to pie.

For those of us who have just started learning, we should basically understand what data should be displayed with what graphics, so let's summarize it together.

Here are 2 ways to visualize Python, namely Seaborn and Pyecharts, these 2 libraries are easy to learn, easy to use, you can quickly draw concise and beautiful charts, and the amount is small, it is very convenient to use, I will briefly introduce the installation and use of these 2 libraries, the experimental environment win10+, the main content is as follows:

Seaborn: This is a visualization library based on matplotlib, which is a more advanced encapsulation of matplotlib, which greatly facilitates our data visualization, saves the configuration of many matplotlib default parameters, ** the amount is small, and the mapping is beautiful, let me briefly introduce this library:

1.To install Seaborn, just enter the command "pipinstallseaborn" in the cmd window, as follows:

2.After the installation is successful, we can conduct a simple test, and the main anti-rotten beats are as follows (official example):

The screenshot of the program running is as follows, and the mapping effect is not bad:

pyecharts: This is a python interface for echarts, with the help of echarts, we can quickly draw concise and beautiful visual charts. Easy to learn, easy to understand, easy to use, very suitable for data visualization, let me briefly introduce this library:

2.After successful installation, we can conduct a simple test, the main ** is as follows:

A screenshot of the program is as follows

At this point, the two python visualization methods have been introduced. In general, these 2 visualization libraries are very convenient to use, easy to learn, easy to use, if you are interested, you can refer to the official website tutorial, try it, of course, you can also use matplotlib, ggplot2 and other visualization libraries, as long as it is suitable for your own project, you can do it, there are also relevant tutorials and materials on the Internet, you can search for it if you are interested, I hope the content shared above can help you, and you are also welcome to comment and leave a message.

Data visualization is one of the most common application areas of Python, data visualization is the process of expressing a set of data graphically with the help of graphical means, and using data analysis and development tools to discover unknown information in it.

There is a saying that is widely circulated in the academic world, a picture worths thousand words, that is, a picture is worth a thousand words. In class, the example I often give is that if you see someone ** when you are browsing the circle of friends, when you see someone ** when the topic is very attractive, we will click into it, maybe the first few paragraphs will be read very seriously, and when the article is very long, it will be followed by ten lines at a glance, and lose interest in reading.

Therefore, the data, ** and text are expressed in the form of charts, which can not only improve the reader's interest in reading, but also intuitively express the content they want to express.

There are many python visualization libraries, and here are a few of the most commonly used ones.

matplotlib

It is the originator of many data visualization libraries in Python, and it is also the most basic third-party database for data visualization, with a simple and easy-to-understand language style, especially suitable for beginners to get started.

seaborn

Seaborn is a more advanced API wrapper on top of MattPlotlib, which makes it easier to make graphs, in most cases Seaborn can make very attractive plots, and with MattplotLib you can make plots with more features. Seaborn should be seen as a complement to matplotlib, not a replacement.

pyecharts

pyecharts is a powerful data visualization tool that combines Python and Echarts, generating exquisite charts and good interactivity, which can be easily integrated into mainstream web frameworks such as Flask, Sanic, and Django, and has been recognized by many developers.

BokehBokeh is an interactive visualization library for web browsers that provides elegant, concise construction of versatile graphics and high-performance interactivity on large or streaming datasets.

Python is a visual library that can easily and efficiently generate a variety of charts, and here are some common ones.

Column chart. Bar chart.

Slope map. Nightingale Rose Figure.

Radar chart. Word cloud diagrams.

Scatter plot. Contour map.

Waterfall chart. Correlation coefficient plot.

Scatter plots.

Histogram. Box plot.

Kernel density estimation plot.

Line chart. Area chart.

Calendar chart. Pie charts.

Donut chart. Mosaic diagram.

Waffle charts. There are other graphs such as geospatial, so I won't list them all, but we'll start with matplotlib, the most commonly used visualization library.