CODEWITHSUNDEEP
pythonpandasdata-visualizationbokehgoogle-sites-2016
Adam Caudell
Adam Caudell

Reputation: 39

Insert a dynamic Bokeh Chart into Google Sites

I am using the Bokeh library in Python to create interactive web charts. I want to store these charts on a Google Site so it is easily accessible within my organization's domain, but I cannot figure out how to keep it live-synced.

The data for the charts is in a Google Spreadsheet. I import that data into a Pandas DataFrame and then create the visualization in Bokeh. I can get the html output file and copy that into an Embed widget in the Google Site, but that remains a static visualization. I want the chart to refresh as new data is added to the google sheet. Is this possible? Here is the same of my code:

from bokeh.plotting import figure
from bokeh.io import output_file, show, save
from bokeh.resources import CDN
from bokeh.embed import file_html
import gspread
from oauth2client.service_account import ServiceAccountCredentials
import pandas as pd

#Connect to Google Sheet and get data

scope = ['https://spreadsheets.google.com/feeds',
         'https://www.googleapis.com/auth/drive']

credentials = ServiceAccountCredentials.from_json_keyfile_name(
         'WUSD Dashboard.json', scope) # Your json file here

gc = gspread.authorize(credentials)

wks = gc.open("Walk Through Data (Responses)").sheet1

data = wks.get_all_values()
headers = data.pop(0)

#Make dataframe from Google Data
#Change data types for numeric columns to numeric data type (they import as a string)

df = pd.DataFrame(data, columns=headers)
df["Engagement (x)"]=pd.to_numeric(df["Engagement (x)"])
df["Evidence of learning (y)"] = pd.to_numeric(df["Evidence of learning (y)"])

cat = df["Site"]
x = df["Engagement (x)"]
y = df["Evidence of learning (y)"]

output_file("Scatter.html")

f = figure()
f.scatter(x,y)

save(f)

How can I trigger the visualization to run the Python Script that will fetch the updated data every time the page loads? Is this even possible?

Upvotes: 0

Views: 586

Answers (1)

Mikel Matticoli
Mikel Matticoli

Reputation: 21

This is a bit late but Google Collab might be a viable option:

https://colab.research.google.com/

You can create a python "notebook" that will run your code on-demand. This can be stored in Google Drive and anyone can access and run it on Google Collab at any time

Upvotes: 2

Related Questions