nonconformist
nonconformist copied to clipboard
Python implementation of the conformal prediction framework.
What's wrong with this code? Given my "y_train" has 2 unique values: 0 and 1. **Output: Unique values in target variable: 2 Classes in IcpClassifier before fit: None Classes in...
Hi, thanks for your work on conformal prediction. Currently I'm working on application to a large (or middle?) dataset (approx. 10k*10k), for which local training seems to never end. Any...
Hi, I have been trying to build the acp model but i am getting the type object 'ClassifierNc' has no attribute 'get_problem_type' error.  Please let me know to overcome...
Hi donlnz, thanks for creating this lib. Did you already plan how to implement the exchangeability testing from: [2] Fedorova, V., Gammerman, A., Nouretdinov, I., & Vovk, V. (2012). Plug-in...
Good afternoon and thank you for the CP library, Do I understand correctly that Mondrian ICP for regression is not implemented here? I see a condition parameter in IcpRegressor class,...
Hello, Thanks for the repository. Are you or other contributors planning to extend the package to allow multi-output regression models?
Hi, Could It be possible implement a function for predict_proba when you are using acp-crossSampler or another ACP methods? I am doing something like this:  I am using a...
Hi, We've been working on a repo that makes using the `nonconformist` package easier and provides some visualization https://github.com/SpikeLab-CL/cpadapter I wonder if it makes sense for some it to be...
Hi, I am trying to apply mondrian conforman with TCP of nonconformist, but when I apply predict I have an Error like ICP doesn't contain _get_stats. I have been reviewing...
Thank you for your library! Is there any easy way to save the fitted and calibrated conformal predictors for re-use? I'd like to make conformal predictions in an online setting....