Replies: 1 comment 4 replies
-
This example is doing horizontal federated learning xgboost, which means that each site will have part of the whole data, while each data sample will need to contain all of the features. If you want to experiment with each of the sites containing different features, that would be supported by vertical federated xgboost (https://github.com/NVIDIA/NVFlare/tree/main/examples/advanced/vertical_xgboost). |
Beta Was this translation helpful? Give feedback.
4 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, I got this error message when I run federated learning xgboost, https://github.com/NVIDIA/NVFlare/tree/2.3/examples/advanced/xgboost/tree-based , with my data. Should the number of feature of site1(client1) and the number of feature of site2 be same? Is it caused because cross-site-validation (i.e., validating the model trained by site1 with site2 data, etc)? If not, what caused this error?
(2612 vs. 3734) : Number of columns does not match number of features in booster,
where 2612 is the number of feature of site1(client1), and 3734 is the number of feature of site2.
linux OS and 3.8 python and NVFlare 2.3 are used. Thank you.
Beta Was this translation helpful? Give feedback.
All reactions