Getting Started with Plotly-Python

8 min readJul 9, 2021


The Plotly Python library is an interactive open-source library. This can be a very helpful tool for data visualization and understanding the data simply and easily. plotly graph objects are a high-level interface to plotly which are easy to use. It can plot various types of graphs and charts like scatter plots, line charts, bar charts, box plots, histograms, pie charts, etc.

So you all must be wondering why plotly over other visualization tools or libraries? Here’s the answer –

  • Plotly has hover tool capabilities that allow us to detect any outliers or anomalies in a large number of data points.
  • It is visually attractive that can be accepted by a wide range of audiences.
  • It allows us for the endless customization of our graphs that makes our plot more meaningful and understandable for others.

Ok, enough theory let’s start.


plotly may be installed using pip:

$ pip install plotly==5.1.0

or conda:

$ conda install -c plotly plotly=5.1.0

This package contains everything you need to write figures to standalone HTML files.

Note: No internet connection, account, or payment is required to use Prior to version 4, this library could operate in either an “online” or “offline” mode. The documentation tended to emphasize the online mode, where graphs get published to the Chart Studio web service. In version 4, all “online” functionality was removed from the plotly package and is now available as the separate, optional, chart-studio package (See below). version 4 is "offline" only, and does not include any functionality for uploading figures or data to cloud services.

import plotly.graph_objects as go
fig = go.Figure(data=go.Bar(y=[2, 3, 1]))
fig.write_html('first_figure.html', auto_open=True)

Plotly chart in Dash

Dash is the best way to build analytical apps in Python using Plotly figures. To run the app below, run pip install dash, click "Download" to get the code and run python

Get started with the official Dash docs and learn how to effortlessly style & deploy apps like this with Dash Enterprise.

JupyterLab Support

For use in JupyterLab, install the jupyterlab and ipywidgets packages using pip:

$ pip install “jupyterlab>=3” “ipywidgets>=7.6”

or conda:

$ conda install “jupyterlab>=3” “ipywidgets>=7.6”

These packages contain everything you need to run JupyterLab…

$ jupyter lab

and display plotly figures inline using the plotly_mimetype renderer...

import plotly.graph_objects as go
fig = go.Figure(data=go.Bar(y=[2, 3, 1]))

or using FigureWidget objects.

import plotly.graph_objects as go
fig = go.FigureWidget(data=go.Bar(y=[2, 3, 1]))

The instructions above apply to JupyterLab 3.x. For JupyterLab 2 or earlier, run the following commands to install the required JupyterLab extensions (note that this will require node to be installed):

# JupyterLab 2.x renderer support jupyter labextension install jupyterlab-plotly@5.1.0 @jupyter-widgets/jupyterlab-manager

Please check out our Troubleshooting guide if you run into any problems with JupyterLab, particularly if you are using multiple python environments inside Jupyter.

See Displaying Figures in Python for more information on the renderers framework, and see Plotly FigureWidget Overview for more information on using FigureWidget.

Jupyter Notebook Support

For use in the classic Jupyter Notebook, install the notebook and ipywidgets packages using pip:

$ pip install “notebook>=5.3” “ipywidgets>=7.5”

or conda:

$ conda install “notebook>=5.3” “ipywidgets>=7.5”

These packages contain everything you need to run a Jupyter notebook…

$ jupyter notebook

and display plotly figures inline using the notebook renderer…

import plotly.graph_objects as go
fig = go.Figure(data=go.Bar(y=[2, 3, 1]))

or using FigureWidget objects.

import plotly.graph_objects as go
fig = go.FigureWidget(data=go.Bar(y=[2, 3, 1]))

See Displaying Figures in Python for more information on the renderers framework, and see Plotly FigureWidget Overview for more information on using FigureWidget.

Static Image Export supports static image export, using the either the kaleido package (recommended, supported as of plotly version 4.9) or the orca command line utility (legacy as of plotly version 4.9).


The kaleido package has no dependencies and can be installed using pip...

$ pip install -U kaleido

or conda.

$ conda install -c plotly python-kaleido


While Kaleido is now the recommended image export approach because it is easier to install and more widely compatible, static image export can also be supported by the legacy orca command line utility and the psutil Python package.

These dependencies can both be installed using conda:

conda install -c plotly plotly-orca==1.3.1 psutil

Or, psutil can be installed using pip...

pip install psutil

and orca can be installed according to the instructions in the orca README.

Extended Geo Support

Some features rely on fairly large geographic shape files. The county choropleth figure factory is one such example. These shape files are distributed as a separate plotly-geo package. This package can be installed using pip...

$ pip install plotly-geo==1.0.0

or conda.

$ conda install -c plotly plotly-geo=1.0.0

See USA County Choropleth Maps in Python for more information on the county choropleth figure factory.

Chart Studio Support

The chart-studio package can be used to upload plotly figures to Plotly's Chart Studio Cloud or On-Prem services. This package can be installed using pip...

$ pip install chart-studio==1.1.0

or conda.

$ conda install -c plotly chart-studio=1.1.0

Note: This package is optional, and if it is not installed it is not possible for figures to be uploaded to the Chart Studio cloud service.

Where to next?

Once you’ve installed, you can use our documentation in three main ways:

  1. You jump right in to examples of how to make basic charts, statistical charts, scientific charts, financial charts, maps, and 3-dimensional charts.
  2. If you prefer to learn about the fundamentals of the library first, you can read about the structure of figures, how to create and update figures, how to display figures, how to theme figures with templates, how to export figures to various formats and about Plotly Express, the high-level API for doing all of the above.
  3. You can check out our exhaustive reference guides: the Python API reference or the Figure Reference

For information on using Python to build web applications containing plotly figures, see the Dash User Guide.

We also encourage you to join the Plotly Community Forum if you want help with anything related to plotly.

What About Dash?

Dash is an open-source framework for building analytical applications, with no Javascript required, and it is tightly integrated with the Plotly graphing library.

Everywhere in this page that you see, you can display the same figure in a Dash application by passing it to the figure argument of the Graph component from the built-in dash_core_components package like this:

import plotly.graph_objects as go # or as px
fig = go.Figure() # or any Plotly Express function e.g.
# fig.add_trace( ... )
# fig.update_layout( ... )
import dash
import dash_core_components as dcc
import dash_html_components as html
app = dash.Dash()
app.layout = html.Div([
app.run_server(debug=True, use_reloader=False) # Turn off reloader if inside Jupyter

Plotly & Cufflinks

Plotly is an open-source and browser-based graphing library which facilitates interactive plotting. The library is available for a number of programming languages such as Python, R, MATLAB, Perl, Julia, Arduino, and REST, among others.

Cufflinks is another library that connects the Pandas data frame with Plotly enabling users to create visualizations directly from Pandas. The library binds the power of Plotly with the flexibility of Pandas for easy plotting.

Visualising Data With Plotly & Cufflinks

Installing Plotly and Cufflinks

To install Plotly and cufflinks type and execute the following commands

pip install plotly
pip install cufflinks


If you are using the Anaconda distribution make sure to activate your conda environment using the conda activate command before executing the above commands. Use conda deactivate to exit the conda environment.

Hands-On Visualisation

For those who have followed the Pandas plotting, this may sound fairly simple as all we need to do is import some libraries and rename the plot function of Pandas as iplot which stands for interactive plot. Let’s look at some examples.

Importing necessary libraries

For iplots, you might want to stick to using Jupyter Notebook as it has great support for it.

import plotly
import cufflinks as cf
import pandas as pd
import numpy as np

#Enabling the offline mode for interactive plotting locally
from plotly.offline import download_plotlyjs,init_notebook_mode,plot,iplot

#To display the plots
%matplotlib inline

Importing The Dataset

Consider a simple dataset of random numbers. You can either create your own download the same sample used in the below examples by clicking here.

data = pd.read_csv("sample_data.csv")

This is what the dataset looks like :

Creating Interactive Plots

Let’s do a simple plot of the numerical features in the dataset which are columns A, B and C.


As you can see in the above image, Plotly provides a set of options to play around with the visualisation. On the top right corner of the generated plot, you can see 10 buttons that offer different functionalities. In the order of buttons from going left to right we have one to download the plot as a png image, toggle buttons to switch between zoom mode and pan mode, a zoom in and zoom out button, an autoscale button to reset the scale in the plot, a reset axis button, a button to toggle the spike lines in the graph, and pair of toggle button to switch between different mouse hover modes.

On the bottom right corner of the image, we also get an option to export the graph to Plotly’s online gallery.

Like the Pandas plot function, iplot enables users to use the ‘kind’ keyword argument to plot different kinds of plots such as scatter plots, bar plots, histograms etc. Let’s look at some examples.

Scatter plot :

data[['A','B','C']].iplot(kind = 'scatter' , mode = 'markers')

  • kind = “scatter” specifies that we need a scatter plot.
  • mode = “markers” sets the markers(Otherwise the plot will use lines by default).

You can also pass in the axis variables x and y as arguments, for example, x = ‘A’ sets column A along the X axis.

There is a wide range of arguments that can be passed to the iplot method to customise the visualisations. To understand the full list of arguments click here to head to the official documentation of Pyplot for Python,

Bar Graph:

data.iplot(kind = 'bar', x = 'D', y = 'C')

Surface Plots:

Plotly also allows users to plot 3D interactive plots.

data[['A','B','C']].iplot(kind = 'surface', colorscale = 'blues')

The above code displays the numerical features in the dataset along x,y and z-axis in a 3 Dimensional plane.


That was a mere gentle introduction to Pyplot and its capabilities don’t stop here. With almost all sorts of customisations available for each kind of plot, there is no limit for creativity with Pyplot. Pyplot can interactively plot all kinds of charts that are available with Pandas plus many more beautiful visualisations.