Turn complex and difficult data into at-a-glance graphs and charts.

This book demonstrates how to visualize data using Plotly, an interactive visualization package, using R and Python, using employment rate and COVID-19 data as examples.
Part 1 introduces the concept and importance of data visualization, and Part 2 explains the basic concepts of Plotly and essential features such as layout and subplots.
Part 3 introduces Plotly visualization methods and special features for different purposes, such as scatter plots and waterfall charts.
In Part 4, we will learn how to create and implement controls such as buttons and sliders to effectively utilize Plotly visualizations, as well as how to distribute the results online and offline.
Finally, the appendix covers the process of designing and implementing a dashboard using the Dash and Shiny packages to maximize Plotly's capabilities.

Data visualization is the process of making the characteristics of a large amount of data visible at a glance.
So, what can data visualization express about data, and what characteristics can it reveal? We've been learning about data visualization since elementary school.
Even now, fourth graders learn how to represent data using bar graphs in their first semester math textbooks.
That's why the most commonly used method in data visualization is the bar graph, and sometimes line graphs and pie charts are also commonly used.
Before doing data visualization, you must first decide what you want to express with the data visualization and which data visualization method to use.
Data visualization is generally effective in representing the distribution, comparison, trend, composition, correlation, and geography of data.

--- p.10

The Plotly package provides packages bound to each language so that it can be used in various languages ​​such as R, Python, Julia, JavaScript, F#, and Matlab.
Plotly provides a variety of charts and plots for data visualization, including basic charts such as scatter plots and line graphs, statistical charts such as box plots and histograms, scientific charts such as heat maps and ternary plots, and financial charts such as time series charts and candlestick charts.

--- p.20

Typically, a histogram represents the number of data cases that fall within a range set by the bin size, i.e., the data count.
Plotly also draws a basic histogram as a way to express the number of data instances, but it also provides a way to replace this value with a specific function value.
Plotly provides five histogram functions: count, which displays the number of cases, sum, which displays the sum of values ​​belonging to a bin, avg, which displays the average, min, which displays the smallest value, and max, which displays the largest value.
Histogram functions can be used by setting the function name to the histfunc property of the histogram trace.
If you use a function other than count, you must set the variable that should be applied to each function in the y attribute.

--- p.143

Because Plotly is a dynamic data visualization, it offers a variety of features that react based on user interaction online.
There are several functions here, but the function that provides a smooth interpretation of visualized data, such as data values ​​and variations of those values, is hover.
Hover is not a feature unique to Plotly, but was used extensively in the past under the name of 'tooltip'.
That is, it is a function that interacts with the user's mouse or digital pointing device and informs the user of the location where the device is pointing through a small box.
This hover is also commonly called 'mouse over', 'mouse hover', or 'hover box' and can be used in web browsers using CSS.

--- p.234

Static visualizations created with R's ggplot2 or Python's Matplotlib or seaborn make it difficult to obtain additional data from the visualization itself in addition to the data visualized when creating the graph.
Therefore, there is the inconvenience of having to re-code and create additional data to provide visualization.
In particular, recoding is required to check data values ​​at specific locations or zoom in on data from specific sections, and it is impossible to create countless visualizations by predicting user usage patterns.
On the other hand, dynamic visualizations such as Plotly provide various functions that can be used for data analysis, such as checking characteristic data values, zooming in and out, and displaying only specific data.
In Plotly, the set of icons that contain these functions is called the modbar.

--- p.355

There are two ways to create Plotly in Python: using the Plotly library's plotly.graph_objects module and using the plotly.express module.
The Plotly maker also recommends using graph_objects to understand the creation principles of Plotly and use detailed functions.
However, when creating Plotly using the graph_objects module, the code becomes very long and it becomes very confusing when a parenthesis is missing or displayed incorrectly.
So we provide the express module to make Plotly easier.

--- p.420