I did encounter a few issues, e.g. searching for an object with an age, then deleting the object age produces an error.
Error message
File "/home/appuser/venv/lib/python3.7/site-packages/streamlit/scriptrunner/script_runner.py", line 443, in _run_script exec(code, module.__dict__) File "/app/reparatorai/reparator.py", line 84, in <module> useful_data_age=useful_data[np.abs(useful_data.product_age - int(my_age))<=1]
And it would be really nice to filter the brand drop-down menu by the object type (to avoid having to go through the full list of brands).
thanks James for the feedback,
I have corrected the bug, and the list of brands is filtered by the product type
the idea behind it is to motivate people to go for a reparation if they find that there is great chance it can be fixed (based on previous data),
i will look into cleaning the data a bit more for names spelling errors,
This is awesome. How can items be added? I know somebody with a clothes dryer that can’t handle a full load anymore. Both washing machine nor dryers are listed.
The data being used by the app comes from the Open Repair Alliance dataset and relies on there being sufficient information to make a prediction. There are lots and lots of “computer” type devices - mobile phones, laptops etc. A quick query of the data gives me only 14 washing machines. It is heartening to know that they are all listed as “fixed” or “repairable” though.