plotlyPowerpoint 1.3.7

A library using Plotly and Powerpoint to easily generate slides with plotly charts in them

Plotly & Powerpoint

A library used to create powerpoint slides including plotly charts. Use this to automate powerpoint creation including certiain charts/visualizations.

Getting Started

Below is a quick tutorial taking you through how to setup and use this library. I will use public data to showcase how it works.

Step 1 - Install Dependencies

If you are using juptyer notebooks, you may need to do some configuration in order to view charts propertly. We are using Plotly for all graphing, so please refer to the getting started page to learn how to install everything properly.

Step 2 - Install Package

pip install plotlyPowerpoint

Step 3 - Prepare Your Powerpoint

The main function of this library is meant to generate slides which include a plotly chart. In order to do this, you must prepare a powerpoint template with the proper layout. You can design any layout you desire, but make sure that each element you desire to fill with python is created as a placeholder. You can use a powerpoint template I've included in my source code to make things easy (assets/powerpoint_templates/template.pptx).

In order to do this, open up your desired powerpoint and go into the slide master. Insert a new slide or use a current one already created. I recommend to drag this to be first in the order, this will make your life easier later down the road. Now insert the proper elements and arragne them as you please. For the chart, insert an image placeholder and make the apsect ratio similar to a laptop screen. Most images created are in landscape view but not too wide.

Here is an example of what my template looks like.

Step 4 - Load Library and Prepare Data

If you want to follow along with my tutorial, feel free and copy/paste my code. However, ensure you have pydataset installed before you do.

from pptx import Presentation
import pandas as pd
from pydataset import data
import plotlyPowerpoint as pp

############
## Prepare Data
############

#load datasets
df = data('InsectSprays')
df2 = data("JohnsonJohnson")

#Data transformation
df['m2'] = df['count'] * 1.1
df2['year'] = df2['time'].astype(int)
df2 = df2.groupby(['year']).agg({'JohnsonJohnson': 'mean'}).reset_index()

Step 5 - Find Your Powerpoint Elements

In order for this library to know where to place each element, you must give it each element index. See the following code to learn how to do this:

#load presentation
prs = Presentation("template.pptx")

#set the slide we are after - the first slide in the master layout is at index 0
slide = prs.slides.add_slide(prs.slide_layouts[0])

#print index and name
for shape in slide.placeholders:
    print('%d %s' % (shape.placeholder_format.idx, shape.name))

You will need this information later on.

Step 6 - Set Template

Now that we have our powerpoint template, we need to tell the library where to find this file. Do this with the following:

pp.setTemplate("path_to_template")

Step 7 - Define Charts

Here is where the work is done. Creating slides with charts is done by defining an array of dictionary objects. Each object represents one slide. Everything must be represented as a string except your dataframe object. Additionally, this library automatically groups every dataframe down to the level needed for the specific chart. Therefore, you don't have to create a new dataframe for each chart. As long as your df has the right data in it, this function will do the rest.

As a start, I will define two basic charts:

#define slides
slides = [
    {
        "title": "Executive Summary",
        "description": "This is a test",
        "item-index": {
            "slide": 1,
            "title": 0,
            "description": 11
        },
        "charts": [
            {
                "name": "Sessions by Channel over Time",
                "data": df,
                "type": "line",
                "metrics": [
                    {"name": "count", "prettyName": "Testing", "method": "sum"}
                ],
                "axis": "spray",
                "x-axis-title": "Test",
                "item-index": 10
            },
            {
                "name": "Sessions by Channel",
                "data": df,
                "type": "bar",
                "metrics": [
                    {"name": "count", "prettyName": "Testing", "method": "sum"}
                ],
                "axis": "spray",
                "x-axis-title": "Test",
                "item-index": 11
            }
        ]
    }
]

I will go into more detail on each one of these variables in a later section.

Step 8 - Run Function

#run function
pp.createSlides(slides)

This will output a file called output.pptx. I suggest you do not make this your final file, as each time you run this function you will overwrite the powerpoint. Create a copy of this file and start to create your analysis/report there. From here, you can use output.pptx as a slide library, where you can include or delete any chart you create.