nriddiford / WhatStat

A tool to parse and visualise your WhatsApp group chats

WhatStat

A tool to parse and visualise your WhatsApp group chats.

Click here to run the shiny web app.

Please e-mail me if you are experiencing problems with WhatStat.

How to export your WhatsApp chat log

iPhone

Follow the steps outlined on the WhatsApp FAQ page

1. Open the WhatsApp conversation you want to export
2. Tap the contact's name or group subject in the navigation bar
3. Scroll to the bottom and tap "Export Chat"
4. Select "Attach without Media"
5. Select "Save to Phone" or, if you want to email the log, select the Mail app

Android

Follow the steps outlined on the WhatsApp FAQ page

1. Open the chat for the individual or group
2. Tap the Menu button
3. Select More
4. Select "Email chat"
5. Select "Attach without Media"

Install from GitHub

git clone https://github.com/nriddiford/WhatStat.git

Start an R session, and install package:

library(devtools)
install_github("nriddiford/WhatStat")
library(WhatStat)

Run shiny app locally

launchApp()

Parse your chat log

d <- parseR(in_file = 'chatLogs/Dools_chris_chat 3.txt')

Plotting functions

Plot the post count by sender

senderPosts(chatdf = d)

Plot the time messages are sent

senderTime(chatdf = d)

Plot the messages over time (by month)

senderDate(chatdf = d)

Plot the messages over a year (by month)

 senderDate(chatdf = d, filtYear = 2017)

Plot the sentiments per-user for the top 5 users (by word count) using the loughran method of sentiment analysis

chatSentiments(chatdf = d, top_sender = 5, method='loughran')

Make a word corpus using the tm package

c <- makeCorpus(chatdf = d)

Plot the most commonly used words with a minimum word length of 3

 wordFreq(corpus = c, wordlength = 3)

Plot this as a word cloud

chatCloud(chatdf = d, wordlength = 3)