How to download SIPRI arms transfer data as a CSV file instead of a rich-text file.

SIPRI-arms-transfer.md

The Stockholm International Peace Research Institute's Arms Transfers Database provides a tool to download the data. By default, that output is an .rtf rich-text file — not so easy to analyze with your favorite spreadsheet or statistics software. Luckily, getting a CSV of the data isn't very difficult. Here's how.

To get all transfers for 2014, by seller, run this command in your terminal:

curl http://armstrade.sipri.org/armstrade/html/export_trade_register.php --compressed \
    --data 'low_year=2014' \
    --data 'high_year=2014' \
    --data 'seller_country_code=' \
    --data 'buyer_country_code=' \
    --data 'armament_category_id=any' \
    --data 'buyers_or_sellers=sellers' \
    --data 'filetype=csv' \
    --data 'include_open_deals=on' \
    --data 'sum_deliveries=on' \
    --data 'Submit4=Download' \
> sipri-arms-by-seller-2014.csv

• To adjust the timespan, change the low_year and/or high_year variables. E.g., 'low_year=2010'.
• To get data only for specific countries, add their three-letter codes to the seller_country_code and/or buyer_country_code variables, with each country code separated by a space. E.g., focus on U.S. and Canadian sales, use 'seller_country_code=USA CAN'.

Hi @BichTran91, the tool used in the gist above is curl, a command-line utility for fetching HTTP(S) resources: https://en.wikipedia.org/wiki/CURL

Any HTTP library or tool should be able to do the same.

Hi @jsvine, thanks for your reply. I realized that your code runs on Linux (I'm using Windows). I figured out how to do the same using Request in Python. However, the columns I got were 'tidn', 'buyercod', 'sellercod', 'odat', 'odai', 'onum', 'onai', 'ldat', 'term', 'desig2', 'wcat', 'desc', 'coprod', 'nrdel', 'nrdelai', 'delyears', 'buyer', 'seller', 'status', 'tivunit', 'tivorder', 'tivdel'.
Did you encounter this problem? How to get the columns' full names and units?

Hi @BichTran91, unfortunately it has been a long time since I've examined the data or tried to download it. I would recommend examining the SIPRI website (or contacting SIPRI if the website doesn't provide enough information) to get official answers to these questions.

If there is a coding solution for that, I would like to know. Otherwise I agree with you suggestion. Thanks again, @jsvine!

Hi just FYI I just used the code works perfectly! (using mac os 10.14)

thanks a ton, saved a lot of time, greetings from italy (: