This is the coding component of the Data Visualization with Plotly for Python workshop.\ Please find the workshop slides here

Getting Started

Workshop Best Practices

• Copy this notebook so you could edit on your own.
• Make sure your notebook is connected.
• This will be a live coding session. Feel free to edit the code on your copy.
• How to run? Click on the little arrow besides each chunk of code, or click on 'Runtime' - 'Run Selection'.
  • Run the code chunk-by-chunk, don't run all at once.
  • Some codes are dependent on previous codes, if you hit an error, check if you have ran the previous codes.
• If you feel the pace is too fast, please let the instructor know. Otherwise, save your questions till the end of each section.

If you have not installed plotly yet, run the following:

pip install plotly

Requirement already satisfied: plotly in ./opt/anaconda3/lib/python3.8/site-packages (5.3.1)
Requirement already satisfied: tenacity>=6.2.0 in ./opt/anaconda3/lib/python3.8/site-packages (from plotly) (8.0.1)
Requirement already satisfied: six in ./opt/anaconda3/lib/python3.8/site-packages (from plotly) (1.15.0)
Note: you may need to restart the kernel to use updated packages.

Task 0: Importing Libraries

1. The plotly.express module is the recommended entry point, for it is a higher-level module compared to plotly.graph_objects.

import plotly.express as px
type(px)

module

1. To preview the dataset, we will be trying out using the DataFrame.head() function from Pandas vs the create_table function from plotly.figure_factory, which is a graphing module that has even more unique chart types yet included in plotly.js.

import pandas as pd
type(pd)

module

from plotly.figure_factory import create_table
type(create_table)

function

Task 1: Loading Data

The data we are using today is the built-in demo data of Plotly.express module.\ The demo data is from Gapminder, detailing the life expectancy, population, GDP change every five years from 1952 to 2007.\ As you may know, in a data science life cycle, up to 80% of the time may be devoted to cleaning and processing data. The demo data we are using today has been cleaned and is ready for use.\ The main goal of today is to explore the different functionalities of the Plotly library, and this cleaned data will serve our purpose well.

# define our data
gapminder = px.data.gapminder()

Let's preview our data.

1. Using our old friend, Pandas.

pd.DataFrame.head(gapminder)

	country	continent	year	lifeExp	pop	gdpPercap	iso_alpha	iso_num
0	Afghanistan	Asia	1952	28.801	8425333	779.445314	AFG	4
1	Afghanistan	Asia	1957	30.332	9240934	820.853030	AFG	4
2	Afghanistan	Asia	1962	31.997	10267083	853.100710	AFG	4
3	Afghanistan	Asia	1967	34.020	11537966	836.197138	AFG	4
4	Afghanistan	Asia	1972	36.088	13079460	739.981106	AFG	4

1. Using plotly.figure_factory (also known as "ff" by abbreviation) module.

table = create_table(gapminder.head(5)) #display the first 10 rows
table

Task 2: Visualize Population Change with Bar Chart

data_china = gapminder.query("country == 'China'")
print(data_china)
# Let's visualize population (variable name: "pop") using bar chart!
# We will use px.bar(dataset, x=[x variable name], y=[y variable name name])

    country continent  year   lifeExp         pop    gdpPercap iso_alpha  \
288   China      Asia  1952  44.00000   556263527   400.448611       CHN   
289   China      Asia  1957  50.54896   637408000   575.987001       CHN   
290   China      Asia  1962  44.50136   665770000   487.674018       CHN   
291   China      Asia  1967  58.38112   754550000   612.705693       CHN   
292   China      Asia  1972  63.11888   862030000   676.900092       CHN   
293   China      Asia  1977  63.96736   943455000   741.237470       CHN   
294   China      Asia  1982  65.52500  1000281000   962.421381       CHN   
295   China      Asia  1987  67.27400  1084035000  1378.904018       CHN   
296   China      Asia  1992  68.69000  1164970000  1655.784158       CHN   
297   China      Asia  1997  70.42600  1230075000  2289.234136       CHN   
298   China      Asia  2002  72.02800  1280400000  3119.280896       CHN   
299   China      Asia  2007  72.96100  1318683096  4959.114854       CHN   

     iso_num  
288      156  
289      156  
290      156  
291      156  
292      156  
293      156  
294      156  
295      156  
296      156  
297      156  
298      156  
299      156  

bar = px.bar(data_china, x = "year", y = "pop")
bar.show()

Next, let's add some more functionalities to our graph

• Color scale (color): we can use a color scale to indicate life expectancy ("lifeExp").\ There are different built-in continous color scales in Plotly Express.
• Hover label (hover_data): The hover label by default only shows x, y data. If we want to add more, we can pass in different data variables.
• Change data label (labels): Since some data name is abbreviated (e.g. "pop"), we would want to spell it out in the hover label, so to clarify their meaning.

bar = px.bar(data_china, x = "year", y = "pop", color = "lifeExp", color_continuous_scale = "Earth")
bar.show()

bar = px.bar(data_china, x = "year", y = "pop", color = "lifeExp", color_continuous_scale = "Magenta")
bar.show()

Task 3: Visualize GDP and Life Expectancy Data with Scatter Plot

# let's say we want to take a snapshot of the world in 2007
# and observe if the GDP influences the life expectancy
gapminder2007 = gapminder.query("year == 2007") #2007 is a integer value, so it does not be enclosed in another pair of quotes

# make a basic scatter plot with px.scatter(data, x, y)
# life expectancy is "lifeExp", GDP per capita is "gdpPercap"
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp")
scatter.show()

# let's color the data points by continents using the "color" attribute
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp", color = "continent", hover_data = ['country', 'year'])
scatter.show()

Task 4: Customize Interactive Bubble Charts according to Population Size

# let's say we want to advance our scatter plot into a bubble chart
# and we want the size of the data points to change according to population size
# we would want to pass in "size" attribute
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp", 
                     color = "continent", hover_data = ['country', 'year'],
                     size = "pop")
scatter.show()

# what do you notice about the bubbles?
# size_max is default 20.
# Should we change something?
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp", 
                     color = "continent", hover_data = ['country', 'year'],
                     size = "pop", size_max = 60)
scatter.show()

Full documentation on Scatter Plot in plotly.express, which has size_max info.

# we need to know what countries are they
# so to help us get insights for outliers
# let's add in another attribute called "hover_name", which will be an overarching title for the hover label.
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp", 
                     color = "continent", hover_data = ['year'],
                     size = "pop", size_max = 60,
                     hover_name = 'country')
scatter.show()

Task 5: Create Facet Plots by Continent

Facet charts can be used to split a multi-trace data visualization into a matrix of data panels.\ In Plotly.express, if you want to facet a larger plot into multiple smaller plots, you simply pass in the variable you want to facet by into an attribute called "facet_col".

# copy the previous code, add in "facet_col" attribute
# isolate by continent
# setting legends
# log x-axis: transformation of graphs
scatter = px.scatter(gapminder2007, x = "gdpPercap", y = "lifeExp", 
                     color = "continent", hover_data = ['year'],
                     size = "pop", size_max = 60,
                     hover_name = 'country',
                     facet_col = "continent",
                     log_x = True)
scatter.show()

Interpretation Challenge: What insights can you gain from this graph?

Write it here:

Task 6: Add Animation to Facet Plots

Let's create animation so that the we can observe the yearly change of GDP VS Life Expectancy data. Code the following:

• "animation_frame": by which variable do you want to observe in every frame of that animation
• "animation_group": sets which data point match which ones across the animation (i.e. each data point represents a country)
• specify range of x and y axis, range_x=[100,100000], range_y=[25,90]
• "labels": rename abbreviated labels

# start by copying the previous code
# change data attribtue to entire dataset, "gapminder"
scatter = px.scatter(gapminder, x = "gdpPercap", y = "lifeExp", 
                     color = "continent", hover_data = ['year'],
                     size = "pop", size_max = 60,
                     hover_name = 'country',
                     facet_col = "continent",
                     log_x = True,
                     animation_frame = "year",
                     animation_group = 'country',
                     range_y = [25, 90],
                     labels = {"pop": "Population",
                               "lifeExp": "Life Expectancy"}
                     )
scatter.show()

Task 7: Represent Geographic Data as Animated Maps

Choropleths are thematic maps in which areas are shaded/ patterned in proportion to the measurement of the displayed statistical variable.\ We are going to shade our maps by lifeExp using our entire dataset.

What is "locations" ='iso_alpha'?

• ISO 3166-1 alpha-2 codes are two-letter country codes defined in ISO 3166-1
• In other words, it's a convention to specify locations
• More on locations here, there are other attributes that define location (e.g. geojson).

# add in 'location'
# 'color', "hover_name", "animation_frame", "projection"
# "color_continuous_scale"
choro = px.choropleth(gapminder, locations = "iso_alpha", color = "lifeExp",
                      hover_name = "country", animation_frame = "year", 
                      projection = "natural earth", color_continuous_scale = "Viridis")

choro.show()
# add in other hover info with "hover_data"
# "color_continuous_scale": because we want the scale to be continuous rather than discrete
# sequential = continuous

# try changing a color scale (e.g. Viridis)
# try changing the "projection" to orthographic

Task 8: Line Plots VS Area Plots

fig = px.line(gapminder, x="year", y="lifeExp", color="continent", line_group="country", hover_name="country",
        line_shape="spline", render_mode="svg")
fig.show()

fig = px.area(gapminder, x="year", y="lifeExp", color="continent", line_group="country", hover_name="country")
fig.show()

Got Questions?\ Ask the Data Services Team at NYU Shanghai Library (shanghai.library@nyu.edu)

Ending Credits\ Tutorial framework: Plotly for Python documentation\ Reference: Kaggle\ Modified and Organized by: Pamela Pan (pp1813@nyu.edu)