[Python] add start_iteration to python predict interface (#3058)

[shiyu1994]

Suggested by Issue #3058. Add start_iteration for python prediction.

[StrikerRUS]

@shiyu1994 Thank you very much! Just some minor comments regarding Python part.

Also, I believe sklearn wrapper should be updated with explicit parameter:

LightGBM/python-package/lightgbm/sklearn.py

Line 615 in a9f5654

def predict(self, X, raw_score=False, num_iteration=None,

LightGBM/python-package/lightgbm/sklearn.py

Line 835 in a9f5654

def predict(self, X, raw_score=False, num_iteration=None,

LightGBM/python-package/lightgbm/sklearn.py

Line 848 in a9f5654

def predict_proba(self, X, raw_score=False, num_iteration=None,

[StrikerRUS]

@shiyu1994 Also, please add some tests that checks new param is working.

Letting know @imatiach-msft about changes in SWIG API.

[guolinke]

ping @shiyu1994 for R's test fails:
══ testthat results ═══════════════════════════════════════════════════════════
[ OK: 558 | SKIPPED: 2 | WARNINGS: 0 | FAILED: 7 ]

1. Failure: start_iteration works correctly (@test_Predictor.R#49)
2. Failure: start_iteration works correctly (@test_Predictor.R#50)
3. Failure: train and predict binary classification (@test_basic.R#32)
4. Error: lgb.intereprete works as expected for binary classification (@test_lgb.interprete.R#38)
5. Error: lgb.intereprete works as expected for multiclass classification (@test_lgb.interprete.R#89)
6. Error: lgb.plot.interepretation works as expected for binary classification (@test_lgb.plot.interpretation.R#38)
7. Error: lgb.plot.interepretation works as expected for multiclass classification (@test_lgb.plot.interpretation.R#87)

[guolinke]

@shiyu1994 it seems there are some lint/doc problems.

[shiyu1994]

@jameslamb @StrikerRUS Could you tell me how to update the R documentation in R-package/man folder? It seems that the tests failed due to code/doc mismatch. I try to use roxygen2 and run roxygenise under R-package folder but the following error occurs. Thanks.

[guolinke]

@shiyu1994
can you try to bulid R package first, then run roxygen2?
https://github.com/microsoft/LightGBM/tree/master/R-package#install

[shiyu1994]

It seems that roxygenise(load = 'installed') works.

[guolinke]

I think .Rdata should be removed.

[guolinke]

also

The downloaded source packages are in
	‘/tmp/Rtmp3IyK3E/downloaded_packages’
Writing NAMESPACE
Writing NAMESPACE
Writing predict.lgb.Booster.Rd
Some R documentation files have changed. Please re-generate them and commit those changes.

    Rscript build_r.R
    Rscript -e "roxygen2::roxygenize('R-package/', load = 'installed')"

[guolinke]

Thank you so much @shiyu1994
@jameslamb @StrikerRUS I am going to merge this. If there are any problems, we can create another PR to fix it.

[github-actions]

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.