Add linear leaf models to json output (fixes #4186) #4329
Conversation
btrotta
requested review from
chivee,
guolinke,
henry0312,
shiyu1994,
StrikerRUS and
wxchan
as code owners
src/io/tree.cpp
Outdated
| str_buf << "\"tree_structure\":{" << "\"leaf_value\":" << leaf_value_[0] << "}" << '\n'; | ||
| if (is_linear_) { | ||
| str_buf << "\"tree_structure\":{" << "\"leaf_value\":" << leaf_value_[0] << ", " << "\n"; | ||
| str_buf << LinearModelToJSON(0); |
There was a problem hiding this comment.
It seems that the right bracket } is missing
| params['linear_tree'] = True | ||
| train_data = lgb.Dataset(X, label=y) | ||
| bst = lgb.train(params, train_data, num_boost_round=5) | ||
| bst.dump_model(5, 0) |
There was a problem hiding this comment.
Shall we use loads to validate that the output JSON content?
There was a problem hiding this comment.
This is done already in dump_model:
LightGBM/python-package/lightgbm/basic.py
Line 3090 in 36957ed
StrikerRUS
changed the title
| @@ -411,3 +411,18 @@ def test_list_to_1d_numpy(y, dtype): | |||
| result = lgb.basic.list_to_1d_numpy(y, dtype=dtype) | |||
| assert result.size == 10 | |||
| assert result.dtype == dtype | |||
|
|
|||
|
|
|||
| def test_dump_model(): | |||
There was a problem hiding this comment.
Thanks a lot for this enhancement!
Please move this test into test_engine.py file because lgb.train() function is from engine.py module.
Also I think it will be useful to add some asserts into this test. Something like assert 'leaf_coeff' in dumped_model for linear model and assert 'leaf_coeff' not in dumped_model for ordinary one.
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
Co-authored-by: Nikita Titov <nekit94-08@mail.ru>
|
Restarted Appveyor because there was unrelated network issue. |
|
This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this. |
As requested here: #3299 (comment)