Powerful open source Python drawing library

The reason why I have been using matplotlib before is to learn its complex syntax, and the hundreds of hours of time cost have been "sinked" in it. This also resulted in me spending countless late nights searching on StackOverflow for how to “format dates” or “add a second Y-axis”.

But we now have a better choice - such as the easy-to-use, well-documented, and powerful open source Python plotting library Plotly. Today I will give you an in-depth experience and learn how it can draw better charts with super simple (even just one line!) code.

All the code in this article has been open sourced on Github, and all charts are interactive. Please use Jupyter notebook to view.

（Github source code address: https://github.com/WillKoehrsen/Data-Analysis/blob/master/plotly/Plotly Whirlwind Introduction.ipynb）

(Example chart drawn by plotly. Image source: plot.ly)

Plotly Overview

plotly's Python package is an open source code library based on plot.js, and then It is based on d3.js. What we actually use is a library that encapsulates plotly, called cufflinks, which makes it easier for you to use plotly and Pandas data tables to work together.

*Note: Plotly itself is a visualization technology company with several different products and open source toolsets. Plotly's Python library is free to use. In offline mode, you can create an unlimited number of charts. In online mode, because Plotly's sharing service is used, you can only generate and share 25 charts.

All visualization charts in this article were completed in Jupyter Notebook using the plotly cufflinks library in offline mode. After completing the installation using pip install cufflinks plotly, you can use the following code to complete the import in Jupyter:

Single variable distribution: histogram and box plot Chart

Univariate analysis charts are often the standard practice when starting data analysis, and histograms are basically one of the necessary charts for univariate distribution analysis (although it still has some shortcomings).

Take the total number of likes on blog posts as an example (see Github for the original data: https://github.com/WillKoehrsen/Data-Analysis/tree/master/medium) and make a simple interactive histogram :

(df in the code is a standard Pandas dataframe object)

(Interaction created using plotly cufflinks Histogram)

For students who are already used to matplotlib, you only need to type one more letter (change .plot to .iplot) to get an interactive chart that looks more beautiful! Clicking on elements on the image reveals detailed information, zooms in and out, and (we’ll get to that next) highlights features like filtering certain parts of the image.

If you want to draw a stacked column chart, you only need to do this:

##Simply perform a simple operation on the pandas data table Process and generate a bar chart:

As shown above, we can integrate the capabilities of plotly cufflinks and pandas in Together. For example, we can first use .pivot() to perform pivot table analysis and then generate a bar chart.

For example, count the number of new fans brought by each article in different publishing channels:

The benefit of interactive charts is that we can explore the data and split sub-items for analysis at will. Box plots can provide a lot of information, but if you can't see the specific values, you're likely to miss a lot of it!

Scatter plot

Scatter plot is the core content of most analysis. It allows us to see the change of one variable over time, or two (or more ) changes in the relationship between variables.

Time Series Analysis

In the real world, a considerable part of the data has time elements. Fortunately, plotly cufflinks comes with built-in functionality to support time series visual analysis.

Taking the article data I published on the "Towards Data Science" website as an example, let us build a data set using the publication time as the index to see how the popularity of the article changes:

In the above picture, we have accomplished several things with one line of code:

• Automatically generate beautiful time series X-axis
• Add a second Y-axis because the ranges of the two variables are not consistent
• Put the article title in the label displayed when the mouse is hovered

In order to display For more data, we can easily add text annotations:

(Scatter plot with text annotations)

In the code below, we color a two-variable scatter plot by the third categorical variable:

Next We're going to get a little more complicated: logarithmic axes. We achieve this by specifying the layout parameter of plotly (for different layouts, please refer to the official document https://plot.ly/python/reference/), and we combine the size of the point (size parameter) and a The numerical variable read_ratio (reading ratio) is bound. The larger the number, the larger the size of the bubble.

If we want to be more complicated (see Github source code for details), we can even stuff 4 into one picture A variable! (However, it is not recommended that you actually do this)

As before, we can combine pandas and plotly cufflinks to achieve many useful charts:

# It is recommended that you check the official documentation or source code, which contains more examples and function instances. With just one or two lines of code, you can add useful elements such as text annotations, auxiliary lines, and best-fit lines to your chart, while maintaining the original interactive functions.

Advanced drawing functions

Next, we will introduce several special charts in detail. You may not use them very often, but I guarantee that as long as you use them well, It will definitely impress people. We are going to use plotly’s figure_factory module, which can generate awesome charts with just one line of code!

Scatter plot matrix

If we want to explore the relationship between many different variables, the scatter plot matrix (also known as SPLOM) is a great choice:

#Even such complex graphics are fully interactive, allowing us to explore the data in more detail.

Relationship heat map

In order to reflect the relationship between multiple numerical variables, we can calculate their correlation and then visualize it in the form of a labeled heat map:

Customized Theme

In addition to the endless variety of charts, Cufflinks also provides many different coloring themes, allowing you to easily switch between different chart style. The following two pictures are the "space" theme and the "ggplot" theme respectively:

##In addition, there are 3D charts (surface and bubble Bubble):

For users who are interested in research, it is not difficult to make a pie chart:

Editing in Plotly Chart Studio

After you generate these charts in Jupyter Notebook, you will find that the lower right corner of the chart appears A small link that says "Export to plot.ly". If you click on this link, you will jump to a "Chart Workshop" (https://plot.ly/create/).

Here you can further revise and polish your diagram before final presentation. You can add annotations, choose the colors of certain elements, organize everything and produce an awesome diagram. Later, you can also publish it on the web, generating a link for others to review.

The following two pictures were made in the chart workshop:

Having said so much, you can read them all Are you tired of watching it? However, we have not exhausted all the capabilities of this library. Due to space limitations, there are some better charts and examples, so please visit the official documents of plotly and cufflinks to view them one by one.

(Plotly interactive map showing wind farm data across the United States. Source: plot.ly)

Finally...

The worst thing about the sunk cost fallacy is that people often only realize how much time they have wasted when they give up their previous efforts.

When choosing a drawing library, the functions you need most are:

One line of code chart required to quickly explore data

Split /Interactive elements needed to study data
• Options to drill down into details when needed
• Easy to customize before final presentation

From now on , the best choice to use Python language to achieve the above functions is plotly. It allows us to quickly generate visual diagrams, and interactive features allow us to better understand the information.

I admit that plotting is definitely the most enjoyable part of data science work, and plotly makes completing these tasks even more enjoyable.

(Use a graph to show how the joy of drawing with Python changes over time. Source towardwardsdatascience.com)

2022 is the time to upgrade you Python drawing library, let yourself become faster, stronger and more beautiful in data science and visualization!

The above is the detailed content of Powerful open source Python drawing library. For more information, please follow other related articles on the PHP Chinese website!