Plotly is one of the most powerful and interactive Python Data Visualization libraries . Here, interactive is the main keyword that distinct Plotly from other Python Data Visualization libraries. In this Python tutorial, I will walk you through the different types of graphs that you can plot using the Python Plotly Library.
When it comes to Data Science with Python we should always use the Jupyter Notebook to code because it is specially designed for Data Science related work, and Python Data Science libraries work very efficiently with Jupyter Notebook. If Jupyter Notebook is not Installed in your System, please check this article on How to Install Jupiter notebook for Python .
Install Required Libraries
For this tutorial, we will be using some of Python's most popular Data Science libraries along with Plotly. So make sure that all those libraries are installed for your Python environment. You can install all the below-mentioned libraries using the Python pip command on the terminal/Command prompt or directly use the Jupyter Notebook to install libraries .
plotly
pip install plotly
pandas
pip install pandas
numpy
pip install numpy
cufflinks (API connect Pandas Data frame with Plotly)
pip install cufflinks
Plotly Basic
Now launch your Jupyter notebook and start with importing the required modules.
import numpy as np
import pandas as pd
import plotly.express as px
import cufflinks as cf #to plot pd data frame
%matplotlib inline
cf.go_offline()
Now let's first create a random numpy array of 20 rows and 5 columns and then convert it into a Pandas DataFrame. Then using the cufflinks API, we can call the
iplot()
method on the DataFrame to plot the DataFrame values on Plotly.
arr = np.random.randn(20,5) #random 2d array
df = pd.DataFrame(arr, columns=['a', 'b','c','d','e']) #data frame
df.iplot() #plot plotly on data frame
When you execute it on the Jupyter notebook, you will see the similar output
As you can see that Plotly give more option than the standard Python matplotlib library and its graph are more interactive.
Plot Line Graph with Plotly
Now let's plot a line Graph with Plotly. We will start with importing the required libraries.
import plotly.express as px
Now we require a data set on which we can Plot the graph. Luckily, the Plotly library comes with some built-in custom data sets for example purposes. You can check out all the data sets from the Plotly Data Package official website.
Here we will be using the Medel for Olympic Short Track Speed Skating data set provided by Plotly as
data.medals_wide()
#https://plotly.com/python-api-reference/generated/plotly.data.html#plotly.data.medals_wide
df_medals = px.data.medals_wide() #
#print(df_medals) #to see the data frame
#plot line graph
px.line(df_medals, x='nation', y='gold', title="Olympic Medals")
Execute
Plot Bar Charts with Plotly
Bar charts can be used to visualize the result, which varies in the time interval. To plot a bar chart using Plotly, we use the Plotly
bar()
method. Here we will be Plotting the population bar graph for India. For the data set, we will be using the Plotly data package
gapminder()
function. Which will return a data frame of countries with their population, life experience, GDP, etc., details.
But here, we are only interested in a specific country, "India," so we will be performing a query on the data set and retrieve only India detail. To perform the query on the data set, we use the query() method.
import plotly.express as px
df_world = px.data.gapminder()
df_india = df_world.query("country =='India'")
#plot bar chart
px.bar(df_india, x='year', y='pop')
Execute
Plot Scatters Plot with Plotly
Scatter Plot is a graphical dot value representation for two or more two variables. To plot the scatter plot in Plotly, we use the Plotly
scatter()
method. For this tutorial, we will use the Plotly Data Package
tips()
data set and plot a graph for tips paid by the men and women on the Total bill paid.
import plotly.express as px
df_tips = px.data.tips()
#plot scatter plot
px.scatter(df_tips, x="total_bill", y='tip', color='sex')
Execute
Plot Pie Charts with Plotly
The pie chart is represented like a circular piece where every variable is defined as a slice from the piece. Pie charts come very useful when we want to show the individual part or proportion occupied by every variable. To plot a Pie Chart in Plotly, we use the
pie()
method. In the example below, we will be plotting a pie chart representing the population of the top 7 countries of Asia in 2002.
import plotly.express as px
asia_df = px.data.gapminder().query("year== '2002'").query("continent=='Asia'")
#Asia population pie chart for 7 counties
px.pie(asia_df[:7], values="pop", names="country", title="Asia Population in 2002")
Execute
Plot Histogram with Plotly
Histogram plots are similar looking to the bar plots, but the histogram is used to visualize the data distribution. With the histogram, we can easily estimate the major surge of values. For the example below, we will create two random arrays of 7000 values representing dice numbers between 1 to 6 and sum them. And with a histogram, we will see which number is occurred most of the time. To plot a Histogram in Plotly, we use the Plotly
histogram()
method.
import numpy as np
dice_roll_1 = np.random.randint(1, 7, 7000)
dice_roll_2 = np.random.randint(1, 7, 7000)
dice_sum = dice_roll_1 + dice_roll_2
#plot histogram
px.histogram(dice_sum, nbins = 6, labels={'value':"Dice Roll"}, title="Roll Dice 7000 times")
Execute:
Conclusion
This tutorial is all about Plotly basic and how to plot basic plots using the Python Plotly library. The Python Plotly library is capable of doing more stuff. We will recommend you go through Plotly official documentation , where they have provided the best tutorials and exexamples to get started with Plotly.
People are also reading:
Leave a Comment on this Post