This script uses the excellent instaloader to scrape data for a given list of usernames and exports the results as two CSV files and a graph file:
- One CSV file (
accounts.csv) containing one row per scraped username, with metadata such as biography, number of followers, number of posts, et cetera - One CSV file (
posts.csv) containing one row per post, with metadata such as post URL, caption, hashtags, likes, et cetera - One Gephi-compatible GDF file (
follower-network.gdf) containing a graph representing followers and followees for each account. This is a directed graph with accounts as nodes, and directed edges representing follower/followee relationships. If accountinternetpersonfollows accountwebuser, an edge frominternetpersontowebuseris created.
These files will, by default, be created in the folder you run the script from.
This script requires Python 3. It also requires instaloader to work. You can easily install it with pip from the script's directory:
pip install -r requirements.txt
You may need to use pip3 if pip does not work or uses Python 2 (you can check this by running pip -V). Using this
command will ensure the instaloader version required by the script is installed.
Due to Instagram's limitations, the scrape needs to be done 'from' a logged-in account. To this end, you need to create
a file config.py in the script's folder containing your Instagram username and password. You can use
config.py-example as a template.
Once you have created this file, run the script this way:
python scrape-batch.py --batchfile batch.txt
This will scrape all accounts listed in batch.txt, assuming one account per line. If you want the result files to be
created elsewhere, you can pass an additional argument:
python scrape-batch.py --batchfile batch.txt --output /home/sam/scrape-results
This will create the files in the folder /home/sam/scrape-results. The folder needs to exist.
This script was developed by the Digital Methods Initiative and is licensed with the MIT License. See LICENSE for more information.