It would be very useful to be able to download the actual data collected for creating a roast profile graph in CSV or similar data format. This would facilitate further plotting and analysis using other tools like Excel, or a stats package. Thank you for your consideration.
1 Like
A while ago @john_l gave me access to a Python script he wrote that would dump the data into csv. Maybe he can help you out here 
I’m not a coder so I was very appreciative that he provided it to those who asked.
1 Like
A least with RT3, the data is cached on your system already so I worked with the cached data.
The code is here: GitHub - jglogan/roastime-data: Tools for analyzing Aillio Bullet roast data.
2 Likes
@ekasten.y3ZY just to clarify the data you’d get from @john_l 's script are data associated with meta data that is available - not every single reading of the temp probes in order to recreate the RT/RW graphs. To get that you can download each roast’s data individually from RW or RT. John’s script is a bulk dump of meta data you want.
1 Like
If you want some (not so pretty) code to get the samples for a roast, look at the Jupyter notebook in the repo.
1 Like
Just go to roastworld and under the actioins button on the right select roast csv.
1 Like
Anyone know where the roast data is stored on the computer? Want to know for future computers so that I can migrate data and back things up even when offline. I dislike work computers to be online, as it almost always results in bugs eventually. Things are always more stable when offline.
The support staff told me once, but I tried to follow their advice and could not locate them.
If you’re on a Win10 machine go to c:/users//AppData/Roaming in there you should see 2 roast time folders (I think) and one of them has folders for your beans and roasts. I’m not on the laptop that I use for roasting so I can’t give you exact info on that last bit there.
See roastime-data/dump_roasts.py at main · jglogan/roastime-data · GitHub
or TL;DR:
- Linux:
${XDG_CONFIG_HOME}/roast-timeif this environment variable is defined; otherwise${HOME}/.config/roast-time. - Windows:
C:\Users\AppData\Roaming\roast-time, as @blacklabs said. - Mac:
${HOME}/Library/Application Support/roast-time.
3 Likes
I also have a python notebook for extracting the raw data from .jsons with a fair bit of clean up, transformation, and new calculations available here: GitHub - rockpyer/bullet-roasting: Notebook for ingesting, expanding, displaying, and analyzing Allio Bullet coffee roasting data.
I expect to work on it a bit more this year, adding more analytics and predictions as I learn how
2 Likes
is this possible to use on a mac? i am only vaguely familiar with python.
Yes. The default directories are all set up for a mac. The easiest way to get started is to install Anaconda Navigator and open it in Jupyter Notebook
1 Like
I developed mine initially for Mac and IIRC @blacklabs verified that it worked on Windows as well. The code is there for it to work on Linux systems too, though I don’t know if anyone has tested it there.
1 Like
you guys have done brilliant work! can’t wait to dive in. love this approach.
Definitely vouch for @john_l 's generosity to share his python code to dump my data in my Win10 environment. I’ve been religiously dumping the data but just effing hadn’t had time to work with it… the down side of also running a business (non-coffee) and client deliverables takes precedence so I can pay my mortgage and feed this coffee hobby! 
1 Like
I also worked on some python code to extract some roast data.
I haven’t touched it in a while because I got it to the point where it got all the info I needed out. I have more plans for it to recreate the plot images and a few other things but life is busy, so it’s on the back-burner.
But it works, if you want more inspiration
2 Likes
Have to admit, haven’t used these yet, and have been going off memory.
But as the roasts stack, up, I have to get something like this going. I cupped a few favorites over the week, and it was amazing how closely aligned my favorite roasts were. Certainly able to trend with the Aillio and get somewhat consistent results, despite the small size.