combo documentation - read the docs · return type return score combo.search.score.pi(predictor,...
TRANSCRIPT
Contents
1 Introduction 31.1 Getting Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2 Citation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.3 Credits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.4 Licence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Methodology 52.1 Generals in Bayesian optimization approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Technical features in COMBO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Installation 73.1 Required Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73.2 Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73.3 Uninstall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 Code Overview 94.1 Modules in COMBO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94.2 Structure of COMBO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94.3 Classes and functions in COMBO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5 Tutorials 235.1 Random search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235.2 Bayesian optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6 Indices and tables 25
Bibliography 27
i
CHAPTER 1
Introduction
COMmon Bayesian Optimization Library (COMBO)
Bayesian optimization [Moc74] has been proven as an effective tool in accelerating scientific discovery. A standardimplementation (e.g., scikit-learn), however, can accommodate only small training data. COMBO [URH+16] is highlyscalable due to an efficient protocol that employs Thompson sampling [CL11], random feature maps [RR08], one-rankCholesky update [GGMS72] and automatic hyperparameter tuning [CER06].
1.1 Getting Help
The latest version of COMBO and documentation can always be found at https://github.com/tsudalab/combo.
1.2 Citation
We ask that you acknowledge the use of COMBO in any publications arising from the use of this code through thefollowing reference
[ref] Tsuyoshi Ueno, Trevor David Rhone, Zhufeng Hou, Teruyasu Mizoguchi and Koji Tsuda, COMBO: An Effi-cient Bayesian Optimization Library for Materials Science, Materials Discovery, (2016), in press. Available fromhttp://dx.doi.org/10.1016/j.md.2016.04.001
Bibtex file for citing COMBO
@article{Ueno2016,title = "COMBO: An Efficient Bayesian Optimization Library for Materials Science ",journal = "Materials Discovery",volume = "",number = "",pages = "-",year = "2016",note = "",doi = "http://dx.doi.org/10.1016/j.md.2016.04.001",url = "http://www.sciencedirect.com/science/article/pii/S2352924516300035",author = "Tsuyoshi Ueno and Trevor David Rhone and Zhufeng Hou and Teruyasu Mizoguchi and Koji Tsuda",}
3
combo Documentation, Release 0.1a
1.3 Credits
1.4 Licence
This package is distributed under the MIT License.
The MIT License (MIT).
Copyright (c) <2015-> <Tsuda Laboratory>
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documen-tation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use,copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whomthe Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of theSoftware.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PAR-TICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHTHOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTIONOF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFT-WARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
1.5 References
4 Chapter 1. Introduction
CHAPTER 2
Methodology
2.1 Generals in Bayesian optimization approach
∫︁ 𝑏
𝑎
𝑓 =
[︂𝑓(𝑎 +
𝑑𝑥
2) + 𝑓(𝑎 + 𝑑𝑥 +
𝑑𝑥
2) + (𝑓(𝑎 + 2𝑑𝑥 +
𝑑𝑥
2) + · · ·
]︂𝑑𝑥
2.1.1 Gaussian process
2.1.2 Acquisition Functions for Bayesian Optimization
2.2 Technical features in COMBO
2.2.1 Thompson sampling [CL11]
2.2.2 Random Feature Map [RR08]
2.2.3 Automatic hyperparameter tuning [CER06]
2.3 References
5
CHAPTER 3
Installation
3.1 Required Packages
• Python 2.7.x
• numpy >=1.10
• scipy >= 0.16
• Cython >= 0.22.1
• mpi4py >= 2.0 (optional)
1. Copy the following contents to a text file named ‘requirements.txt’
## To install these requirements, run## pip install -U -r requirements.txt## (the -U option also upgrades packages; from the second time on,## just run## pip install -r requirements.txt#### NOTE: before running the command above, you need to install a recent version## of pip from the website, and then possibly install/upgrade setuptools using## sudo pip install --upgrade setuptools## numpynumpy >=1.10
## scipyscipy >= 0.16
##Cython >= 0.22.1
## mpi4pympi4py >= 2.0 (optional)
2. Run the command
> pip install -U -r requirements.txt
3.2 Install
1. Download or clone the github repository, e.g.
7
combo Documentation, Release 0.1a
> git clone https://github.com/tsudalab/combo.git
2. Run setup.py install
> cd combo> python setup.py install
3.3 Uninstall
1. Delete all installed files, e.g.
> python setup.py install --record file.txt> cat file.txt | xargs rm -rvf
8 Chapter 3. Installation
CHAPTER 4
Code Overview
4.1 Modules in COMBO
Instruction to import the modules in COMBO:
Relative import 'gp', should be 'combo.gp' (relative-import)Relative import 'opt', should be 'combo.opt' (relative-import)Relative import 'blm', should be 'combo.blm' (relative-import)Relative import 'misc', should be 'combo.misc' (relative-import)Relative import 'search', should be 'combo.search' (relative-import)Relative import 'predictor', should be 'combo.predictor' (relative-import)Relative import 'variable', should be 'combo.variable' (relative-import)
4.2 Structure of COMBO
The files in each directory:
|-- blm| |-- basis| | |-- fourier.py| | `-- __init__.py| |-- core| | |-- __init__.py| | `-- model.py| |-- inf| | |-- exact.py| | `-- __init__.py| |-- __init__.py| |-- lik| | |-- gauss.py| | |-- __init__.py| | |-- linear.py| | `-- _src| | |-- cov.py| | `-- __init__.py| |-- predictor.py| `-- prior| |-- gauss.py| `-- __init__.py|-- gp
9
combo Documentation, Release 0.1a
| |-- core| | |-- __init__.py| | |-- learning.py| | |-- model.py| | `-- prior.py| |-- cov| | |-- gauss.py| | |-- gauss.pyc| | |-- __init__.py| | `-- _src| | |-- enhance_gauss.c| | |-- enhance_gauss.pyx| | |-- __init__.py| | `-- __init__.pyc| |-- inf| | |-- exact.py| | `-- __init__.py| |-- __init__.py| |-- lik| | |-- gauss.py| | `-- __init__.py| |-- mean| | |-- const.py| | |-- __init__.py| | `-- zero.py| `-- predictor.py|-- __init__.py|-- misc| |-- centering.py| |-- gauss_elim.py| |-- __init__.py| |-- set_config.py| `-- _src| |-- cholupdate.c| |-- cholupdate.pyx| |-- diagAB.c| |-- diagAB.pyx| |-- __init__.py| |-- logsumexp.c| |-- logsumexp.pyx| |-- traceAB.c| `-- traceAB.pyx|-- opt| |-- adam.py| `-- __init__.py|-- predictor.py|-- search| |-- call_simulator.py| |-- discrete| | |-- __init__.py| | |-- policy.py| | `-- results.py| |-- __init__.py| |-- score.py| `-- utility.py`-- variable.py
10 Chapter 4. Code Overview
combo Documentation, Release 0.1a
4.3 Classes and functions in COMBO
4.3.1 variable object
COMBO defines a python class called variable to setup and handle the dataset in the training and testing. From apython script, variable object can be created like this:
>>>from combo.variable import variable>>>import numpy as np>>>X= np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])>>>t=np.array([[1], [5], [9]])>>>Z=np.array([[2,6,3,4], [9,6,10,8], [22,10,11,12]])>>>vtest=variable(X,t,Z)>>>print vtest.X
class combo.variable.variable
Xarray_like
tarray_like
Zarray_like
get_subset(index)
delete(num_row)
add(X, t, Z)
delete_X(num_row)
delete_t(num_row)
delete_Z(num_row)
add_X(X)
add_t(t)
add_Z(Z)
save(filename)
load(filename)
class combo.predictor.base_predictor
config
model
fit()
prepare()
delete_stats()
get_basis()
get_post_fmean()
4.3. Classes and functions in COMBO 11
combo Documentation, Release 0.1a
get_post_fcov()
get_post_params()
get_post_samples()
combo.misc.centering(X)standardize the data X and remove the columns with zero value
𝑋1,1𝜎 =𝑋𝑖 − �̄�𝑆
𝜎𝑋,𝑆
Parameters X – input data
Return type return the standardized data for the columns with non-zero values
combo.misc.gauss_elim(L, t)
Parameters
• L –
• t –
Return type return
class combo.misc.set_config.set_config
show()
load()
class combo.misc.set_config.search
show()
load()
class combo.misc.set_config.learning
show()
load()
class combo.misc.set_config.batch
show()
load()
class combo.misc.set_config.online
show()
load()
class combo.misc.set_config.adam
show()
load()
combo.misc.set_config.boolean(str)
12 Chapter 4. Code Overview
combo Documentation, Release 0.1a
Parameters str –
Return type return a boolean value
combo.search.utility.show_search_results(history, N)
Parameters
• history –
• N –
combo.search.utility.show_start_message_multi_search(N, score=None)
Parameters
• N –
• score –
combo.search.utility.show_interactive_mode(simulator, history)
Parameters
• simulator –
• history –
combo.search.utility.is_learning(n, interval)
Parameters
• n –
• interval –
combo.search.call_simulator(simu, action)
Parameters
• simu –
• action –
combo.search.score.EI(predictor, training, test, fmax=None)
Parameters
• predictor –
• training –
• test –
• fmax –
Return type return score
combo.search.score.PI(predictor, training, test, fmax=None)
Parameters
• predictor –
• training –
• test –
• fmax –
Return type return score
4.3. Classes and functions in COMBO 13
combo Documentation, Release 0.1a
combo.search.score.TS(predictor, training, test, alpha=1)
Parameters
• predictor –
• training –
• test –
• alpha –
Return type return score
class combo.search.discrete.policy.policy
set_seed()
delete_actions()
write()
random_search()
bayes_search()
get_score()
get_marginal_score()
get_actions()
get_random_action()
load()
export_predictor()
export_training()
export_history()
_set_predictor()
_init_predictor()
_set_training()
_set_unchosed_actions()
_set_test()
_set_config()
class combo.search.discrete.results.history
write()
export_sequence_best_fx()
export_all_sequence_best_fx()
save()
load()
class combo.opt.adam
14 Chapter 4. Code Overview
combo Documentation, Release 0.1a
set_params()
update()
run()
_set_options()
class combo.gp.predictor.predictor
fit()
get_basis()
get_post_params()
prepare()
delete_stats()
get_post_fmean()
get_post_fcov()
get_post_samples()
get_predict_samples()
class combo.gp.core.learning.batch
run()
one_run()
init_params_search()
class combo.gp.core.learning.online
run()
one_run()
disp_marlik()
init_params_search()
get_one_update()
class combo.gp.core.learning.adam
rest()
get_one_update()
class combo.gp.core.model.model
cat_params()
decomp_params()
set_params()
sub_sampling()
export_blm()
4.3. Classes and functions in COMBO 15
combo Documentation, Release 0.1a
eval_marlik()
get_grad_marlik()
get_params_bound()
prepare()
get_post_fmean()
get_post_fcov()
post_sampling()
predict_sampling()
print_params()
get_cand_params()
fit()
class combo.gp.core.prior
cat_params()
decomp_params()
get_mean()
get_cov()
get_grad_mean()
get_grad_cov()
set_params()
set_mean_params()
set_cov_params()
sampling()
class combo.gp.cov.gauss.gauss
cat_params()
print_params()
prepare()
get_grad()
get_cov()
set_params()
supp_params()
decomp_params()
save()
load()
get_params_bound()
cat_params()
16 Chapter 4. Code Overview
combo Documentation, Release 0.1a
rand_expans()
get_cand_params()
combo.gp.inf.exact.eval_marlik(gp, X, t, params = None)
Parameters
• gp –
• X –
• t –
• params –
Return type return marlik
combo.gp.inf.exact.get_grad_marlik(gp, X, t, params = None)
Parameters
• gp –
• X –
• t –
• params –
Return type return grad_marlik
combo.gp.inf.exact.prepare(gp, X, t, params = None)
Parameters
• gp –
• X –
• t –
• params –
Return type return stats
combo.gp.inf.get_post_fmean(gp, X, t, params = None)
Parameters
• gp –
• X –
• t –
• params –
Return type return G.dot(alpha) + fmu
combo.gp.inf.get_post_fcov(gp, X, t, params = None, diag = True)
Parameters
• gp –
• X –
• t –
• params –
4.3. Classes and functions in COMBO 17
combo Documentation, Release 0.1a
• diag –
Return type post_cov
class combo.gp.lik.gauss.gauss
supp_params()
trans_params()
get_params_bound()
get_cov()
get_grad()
set_params()
get_cand_params()
sampling()
class combo.gp.mean.const.const
supp_params()
get_params_bound()
get_mean()
get_grad()
set_params()
init_params()
get_cand_params()
class combo.blm.predictor.predictor
fit()
prepare()
delete_stats()
get_basis()
get_post_fmean()
get_post_fcov()
get_post_params()
get_post_samples()
get_predict_samples()
update()
class combo.blm.basis.fourier
get_basis()
set_params()
18 Chapter 4. Code Overview
combo Documentation, Release 0.1a
show()
_check_params()
_check_len_params()
class combo.blm.core.model.model
prepare()
update_stats()
get_post_params_mean()
get_post_fmean()
sampling()
post_sampling()
predict_sampling()
get_post_fcov()
_set_options()
_init_prior()
combo.blm.inf.exact.prepare(blm, X, t, Psi = None)
Parameters
• blm –
• X –
• t –
• Psi –
combo.blm.inf.exact.update_stats(blm, x, t, Psi = None)
Parameters
• blm –
• x –
• t –
• Psi –
Return type return ( L, b, alpha )
combo.blm.inf.exact.sampling(blm, w_mu = None, N=1, alpha = 1.0)
Parameters
• blm –
• w_mu –
• N –
• alpha –
Return type return (invLz.transpose() + w_mu).transpose()
combo.blm.inf.exact.get_post_params_mean(blm)
4.3. Classes and functions in COMBO 19
combo Documentation, Release 0.1a
Parameters blm –
Return type return blm.stats[2] * blm.lik.cov.prec
combo.blm.inf.exact.get_post_fmean(blm, X, Psi = None, w = None)
Parameters
• blm –
• X –
• Psi –
• w –
Return type return Psi.dot(w) + blm.lik.linear.bias
combo.blm.inf.exact.get_post_fcov(blm, X, Psi = None, diag = True)
Parameters
• blm –
• X –
• Psi –
• diag –
Return type return fcov
class combo.blm.lik.gauss.gauss
get_cov()
get_prec()
get_basis()
get_mean()
set_params()
set_bias()
sampling()
class combo.blm.lik.linear.linear
get_mean()
set_params()
set_bias()
_init_params()
_init_bias()
class combo.blm.prior.gauss.cov_const
get_cov()
get_prec()
set_params()
20 Chapter 4. Code Overview
combo Documentation, Release 0.1a
_trans_params()
class combo.blm.prior.gauss.gauss
get_mean()
get_cov()
get_prec()
set_params()
_init_cov()
4.3. Classes and functions in COMBO 21
CHAPTER 5
Tutorials
5.1 Random search
test.python
5.1.1 Multiple probe search
This is a normal text paragraph. The next paragraph is a code sample
5.1.2 Interactive mode search
5.2 Bayesian optimization
5.2.1 Multiple probe search
5.2.2 Interactive mode search
23
Bibliography
[CER06] Christopher K. I. Williams Carl Edward Rasmussen. Gaussian Processes for Machine Learn-ing. volume of Adaptive computation and machine learning. MIT Press, edition, 2006. URL:http://www.gaussianprocess.org/gpml/.
[CL11] Olivier Chapelle and Lihong Li. An empirical evaluation of thompson sampling. In J. Shawe-Taylor, R. S.Zemel, P. L. Bartlett, F. Pereira, and K. Q. Weinberger, editors, Advances in Neural Information Processing Sys-tems 24, pages 2249–2257. Curran Associates, Inc., 2011. URL: http://papers.nips.cc/paper/4321-an-empirical-evaluation-of-thompson-sampling.pdf.
[GGMS72] Phillip E. Gill, Gene H. Golub, Walter A. Murray, and Michael A. Saunders. Methods for modify-ing matrix factorizations. Math. Comp., 28:505–535, 1972. URL: http://www.ams.org/journals/mcom/1974-28-126/S0025-5718-1974-0343558-6/, doi:http://dx.doi.org/10.1090/S0025-5718-1974-0343558-6.
[Moc74] Jonas Mockus. On bayesian methods for seeking the extremum. In Proceedings ofthe IFIP Technical Conference, 400–404. London, UK, UK, 1974. Springer-Verlag. URL:http://dl.acm.org/citation.cfm?id=646296.687872.
[RR08] Ali Rahimi and Benjamin Recht. Random features for large-scale kernel machines. In J. C. Platt, D. Koller,Y. Singer, and S. T. Roweis, editors, Advances in Neural Information Processing Systems 20, 1177–1184.Curran Associates, Inc., 2008. URL: http://papers.nips.cc/paper/3182-random-features-for-large-scale-kernel-machines.pdf.
[URH+16] Tsuyoshi Ueno, Trevor David Rhone, Zhufeng Hou, Teruyasu Mizoguchi, and KojiTsuda. Combo: an efficient bayesian optimization library for materials science. Materials Dis-covery, ():–, 2016. URL: http://www.sciencedirect.com/science/article/pii/S2352924516300035,doi:http://dx.doi.org/10.1016/j.md.2016.04.001.
[CER06] Christopher K. I. Williams Carl Edward Rasmussen. Gaussian Processes for Machine Learn-ing. volume of Adaptive computation and machine learning. MIT Press, edition, 2006. URL:http://www.gaussianprocess.org/gpml/.
[CL11] Olivier Chapelle and Lihong Li. An empirical evaluation of thompson sampling. In J. Shawe-Taylor, R. S.Zemel, P. L. Bartlett, F. Pereira, and K. Q. Weinberger, editors, Advances in Neural Information Processing Sys-tems 24, pages 2249–2257. Curran Associates, Inc., 2011. URL: http://papers.nips.cc/paper/4321-an-empirical-evaluation-of-thompson-sampling.pdf.
[GGMS72] Phillip E. Gill, Gene H. Golub, Walter A. Murray, and Michael A. Saunders. Methods for modify-ing matrix factorizations. Math. Comp., 28:505–535, 1972. URL: http://www.ams.org/journals/mcom/1974-28-126/S0025-5718-1974-0343558-6/, doi:http://dx.doi.org/10.1090/S0025-5718-1974-0343558-6.
27
combo Documentation, Release 0.1a
[Moc74] Jonas Mockus. On bayesian methods for seeking the extremum. In Proceedings ofthe IFIP Technical Conference, 400–404. London, UK, UK, 1974. Springer-Verlag. URL:http://dl.acm.org/citation.cfm?id=646296.687872.
[RR08] Ali Rahimi and Benjamin Recht. Random features for large-scale kernel machines. In J. C. Platt, D. Koller,Y. Singer, and S. T. Roweis, editors, Advances in Neural Information Processing Systems 20, 1177–1184.Curran Associates, Inc., 2008. URL: http://papers.nips.cc/paper/3182-random-features-for-large-scale-kernel-machines.pdf.
[URH+16] Tsuyoshi Ueno, Trevor David Rhone, Zhufeng Hou, Teruyasu Mizoguchi, and KojiTsuda. Combo: an efficient bayesian optimization library for materials science. Materials Dis-covery, ():–, 2016. URL: http://www.sciencedirect.com/science/article/pii/S2352924516300035,doi:http://dx.doi.org/10.1016/j.md.2016.04.001.
28 Bibliography
Index
Symbols_check_len_params() (combo.blm.basis.fourier method),
19_check_params() (combo.blm.basis.fourier method), 19_init_bias() (combo.blm.lik.linear.linear method), 20_init_cov() (combo.blm.prior.gauss.gauss method), 21_init_params() (combo.blm.lik.linear.linear method), 20_init_predictor() (combo.search.discrete.policy.policy
method), 14_init_prior() (combo.blm.core.model.model method), 19_set_config() (combo.search.discrete.policy.policy
method), 14_set_options() (combo.blm.core.model.model method),
19_set_options() (combo.opt.adam method), 15_set_predictor() (combo.search.discrete.policy.policy
method), 14_set_test() (combo.search.discrete.policy.policy method),
14_set_training() (combo.search.discrete.policy.policy
method), 14_set_unchosed_actions() (combo.search.discrete.policy.policy
method), 14_trans_params() (combo.blm.prior.gauss.cov_const
method), 20
Aadd() (combo.variable.variable method), 11add_t() (combo.variable.variable method), 11add_X() (combo.variable.variable method), 11add_Z() (combo.variable.variable method), 11array_like (combo.variable.variable attribute), 11
Bbayes_search() (combo.search.discrete.policy.policy
method), 14
Ccat_params() (combo.gp.core.model.model method), 15cat_params() (combo.gp.core.prior method), 16
cat_params() (combo.gp.cov.gauss.gauss method), 16combo.blm.basis.fourier (built-in class), 18combo.blm.core.model.model (built-in class), 19combo.blm.inf.exact.get_post_fcov() (built-in function),
20combo.blm.inf.exact.get_post_fmean() (built-in func-
tion), 20combo.blm.inf.exact.get_post_params_mean() (built-in
function), 19combo.blm.inf.exact.prepare() (built-in function), 19combo.blm.inf.exact.sampling() (built-in function), 19combo.blm.inf.exact.update_stats() (built-in function), 19combo.blm.lik.gauss.gauss (built-in class), 20combo.blm.lik.linear.linear (built-in class), 20combo.blm.predictor.predictor (built-in class), 18combo.blm.prior.gauss.cov_const (built-in class), 20combo.blm.prior.gauss.gauss (built-in class), 21combo.gp.core.learning.adam (built-in class), 15combo.gp.core.learning.batch (built-in class), 15combo.gp.core.learning.online (built-in class), 15combo.gp.core.model.model (built-in class), 15combo.gp.core.prior (built-in class), 16combo.gp.cov.gauss.gauss (built-in class), 16combo.gp.inf.exact.eval_marlik() (built-in function), 17combo.gp.inf.exact.get_grad_marlik() (built-in function),
17combo.gp.inf.exact.prepare() (built-in function), 17combo.gp.inf.get_post_fcov() (built-in function), 17combo.gp.inf.get_post_fmean() (built-in function), 17combo.gp.lik.gauss.gauss (built-in class), 18combo.gp.mean.const.const (built-in class), 18combo.gp.predictor.predictor (built-in class), 15combo.misc.centering() (built-in function), 12combo.misc.gauss_elim() (built-in function), 12combo.misc.set_config.adam (built-in class), 12combo.misc.set_config.batch (built-in class), 12combo.misc.set_config.boolean() (built-in function), 12combo.misc.set_config.learning (built-in class), 12combo.misc.set_config.online (built-in class), 12combo.misc.set_config.search (built-in class), 12combo.misc.set_config.set_config (built-in class), 12
29
combo Documentation, Release 0.1a
combo.opt.adam (built-in class), 14combo.predictor.base_predictor (built-in class), 11combo.search.call_simulator() (built-in function), 13combo.search.discrete.policy.policy (built-in class), 14combo.search.discrete.results.history (built-in class), 14combo.search.score.EI() (built-in function), 13combo.search.score.PI() (built-in function), 13combo.search.score.TS() (built-in function), 13combo.search.utility.is_learning() (built-in function), 13combo.search.utility.show_interactive_mode() (built-in
function), 13combo.search.utility.show_search_results() (built-in
function), 13combo.search.utility.show_start_message_multi_search()
(built-in function), 13combo.variable.variable (built-in class), 11config (combo.predictor.base_predictor attribute), 11
Ddecomp_params() (combo.gp.core.model.model method),
15decomp_params() (combo.gp.core.prior method), 16decomp_params() (combo.gp.cov.gauss.gauss method),
16delete() (combo.variable.variable method), 11delete_actions() (combo.search.discrete.policy.policy
method), 14delete_stats() (combo.blm.predictor.predictor method),
18delete_stats() (combo.gp.predictor.predictor method), 15delete_stats() (combo.predictor.base_predictor method),
11delete_t() (combo.variable.variable method), 11delete_X() (combo.variable.variable method), 11delete_Z() (combo.variable.variable method), 11disp_marlik() (combo.gp.core.learning.online method),
15
Eeval_marlik() (combo.gp.core.model.model method), 15export_all_sequence_best_fx()
(combo.search.discrete.results.history method),14
export_blm() (combo.gp.core.model.model method), 15export_history() (combo.search.discrete.policy.policy
method), 14export_predictor() (combo.search.discrete.policy.policy
method), 14export_sequence_best_fx()
(combo.search.discrete.results.history method),14
export_training() (combo.search.discrete.policy.policymethod), 14
Ffit() (combo.blm.predictor.predictor method), 18fit() (combo.gp.core.model.model method), 16fit() (combo.gp.predictor.predictor method), 15fit() (combo.predictor.base_predictor method), 11
Gget_actions() (combo.search.discrete.policy.policy
method), 14get_basis() (combo.blm.basis.fourier method), 18get_basis() (combo.blm.lik.gauss.gauss method), 20get_basis() (combo.blm.predictor.predictor method), 18get_basis() (combo.gp.predictor.predictor method), 15get_basis() (combo.predictor.base_predictor method), 11get_cand_params() (combo.gp.core.model.model
method), 16get_cand_params() (combo.gp.cov.gauss.gauss method),
17get_cand_params() (combo.gp.lik.gauss.gauss method),
18get_cand_params() (combo.gp.mean.const.const
method), 18get_cov() (combo.blm.lik.gauss.gauss method), 20get_cov() (combo.blm.prior.gauss.cov_const method), 20get_cov() (combo.blm.prior.gauss.gauss method), 21get_cov() (combo.gp.core.prior method), 16get_cov() (combo.gp.cov.gauss.gauss method), 16get_cov() (combo.gp.lik.gauss.gauss method), 18get_grad() (combo.gp.cov.gauss.gauss method), 16get_grad() (combo.gp.lik.gauss.gauss method), 18get_grad() (combo.gp.mean.const.const method), 18get_grad_cov() (combo.gp.core.prior method), 16get_grad_marlik() (combo.gp.core.model.model
method), 16get_grad_mean() (combo.gp.core.prior method), 16get_marginal_score() (combo.search.discrete.policy.policy
method), 14get_mean() (combo.blm.lik.gauss.gauss method), 20get_mean() (combo.blm.lik.linear.linear method), 20get_mean() (combo.blm.prior.gauss.gauss method), 21get_mean() (combo.gp.core.prior method), 16get_mean() (combo.gp.mean.const.const method), 18get_one_update() (combo.gp.core.learning.adam
method), 15get_one_update() (combo.gp.core.learning.online
method), 15get_params_bound() (combo.gp.core.model.model
method), 16get_params_bound() (combo.gp.cov.gauss.gauss
method), 16get_params_bound() (combo.gp.lik.gauss.gauss method),
18get_params_bound() (combo.gp.mean.const.const
method), 18
30 Index
combo Documentation, Release 0.1a
get_post_fcov() (combo.blm.core.model.model method),19
get_post_fcov() (combo.blm.predictor.predictor method),18
get_post_fcov() (combo.gp.core.model.model method),16
get_post_fcov() (combo.gp.predictor.predictor method),15
get_post_fcov() (combo.predictor.base_predictormethod), 11
get_post_fmean() (combo.blm.core.model.modelmethod), 19
get_post_fmean() (combo.blm.predictor.predictormethod), 18
get_post_fmean() (combo.gp.core.model.model method),16
get_post_fmean() (combo.gp.predictor.predictor method),15
get_post_fmean() (combo.predictor.base_predictormethod), 11
get_post_params() (combo.blm.predictor.predictormethod), 18
get_post_params() (combo.gp.predictor.predictormethod), 15
get_post_params() (combo.predictor.base_predictormethod), 12
get_post_params_mean() (combo.blm.core.model.modelmethod), 19
get_post_samples() (combo.blm.predictor.predictormethod), 18
get_post_samples() (combo.gp.predictor.predictormethod), 15
get_post_samples() (combo.predictor.base_predictormethod), 12
get_prec() (combo.blm.lik.gauss.gauss method), 20get_prec() (combo.blm.prior.gauss.cov_const method),
20get_prec() (combo.blm.prior.gauss.gauss method), 21get_predict_samples() (combo.blm.predictor.predictor
method), 18get_predict_samples() (combo.gp.predictor.predictor
method), 15get_random_action() (combo.search.discrete.policy.policy
method), 14get_score() (combo.search.discrete.policy.policy
method), 14get_subset() (combo.variable.variable method), 11
Iinit_params() (combo.gp.mean.const.const method), 18init_params_search() (combo.gp.core.learning.batch
method), 15init_params_search() (combo.gp.core.learning.online
method), 15
Lload() (combo.gp.cov.gauss.gauss method), 16load() (combo.misc.set_config.adam method), 12load() (combo.misc.set_config.batch method), 12load() (combo.misc.set_config.learning method), 12load() (combo.misc.set_config.online method), 12load() (combo.misc.set_config.search method), 12load() (combo.misc.set_config.set_config method), 12load() (combo.search.discrete.policy.policy method), 14load() (combo.search.discrete.results.history method), 14load() (combo.variable.variable method), 11
Mmodel (combo.predictor.base_predictor attribute), 11
Oone_run() (combo.gp.core.learning.batch method), 15one_run() (combo.gp.core.learning.online method), 15
Ppost_sampling() (combo.blm.core.model.model method),
19post_sampling() (combo.gp.core.model.model method),
16predict_sampling() (combo.blm.core.model.model
method), 19predict_sampling() (combo.gp.core.model.model
method), 16prepare() (combo.blm.core.model.model method), 19prepare() (combo.blm.predictor.predictor method), 18prepare() (combo.gp.core.model.model method), 16prepare() (combo.gp.cov.gauss.gauss method), 16prepare() (combo.gp.predictor.predictor method), 15prepare() (combo.predictor.base_predictor method), 11print_params() (combo.gp.core.model.model method), 16print_params() (combo.gp.cov.gauss.gauss method), 16
Rrand_expans() (combo.gp.cov.gauss.gauss method), 16random_search() (combo.search.discrete.policy.policy
method), 14rest() (combo.gp.core.learning.adam method), 15run() (combo.gp.core.learning.batch method), 15run() (combo.gp.core.learning.online method), 15run() (combo.opt.adam method), 15
Ssampling() (combo.blm.core.model.model method), 19sampling() (combo.blm.lik.gauss.gauss method), 20sampling() (combo.gp.core.prior method), 16sampling() (combo.gp.lik.gauss.gauss method), 18save() (combo.gp.cov.gauss.gauss method), 16save() (combo.search.discrete.results.history method), 14
Index 31
combo Documentation, Release 0.1a
save() (combo.variable.variable method), 11set_bias() (combo.blm.lik.gauss.gauss method), 20set_bias() (combo.blm.lik.linear.linear method), 20set_cov_params() (combo.gp.core.prior method), 16set_mean_params() (combo.gp.core.prior method), 16set_params() (combo.blm.basis.fourier method), 18set_params() (combo.blm.lik.gauss.gauss method), 20set_params() (combo.blm.lik.linear.linear method), 20set_params() (combo.blm.prior.gauss.cov_const method),
20set_params() (combo.blm.prior.gauss.gauss method), 21set_params() (combo.gp.core.model.model method), 15set_params() (combo.gp.core.prior method), 16set_params() (combo.gp.cov.gauss.gauss method), 16set_params() (combo.gp.lik.gauss.gauss method), 18set_params() (combo.gp.mean.const.const method), 18set_params() (combo.opt.adam method), 14set_seed() (combo.search.discrete.policy.policy method),
14show() (combo.blm.basis.fourier method), 18show() (combo.misc.set_config.adam method), 12show() (combo.misc.set_config.batch method), 12show() (combo.misc.set_config.learning method), 12show() (combo.misc.set_config.online method), 12show() (combo.misc.set_config.search method), 12show() (combo.misc.set_config.set_config method), 12sub_sampling() (combo.gp.core.model.model method),
15supp_params() (combo.gp.cov.gauss.gauss method), 16supp_params() (combo.gp.lik.gauss.gauss method), 18supp_params() (combo.gp.mean.const.const method), 18
Tt (combo.variable.variable attribute), 11trans_params() (combo.gp.lik.gauss.gauss method), 18
Uupdate() (combo.blm.predictor.predictor method), 18update() (combo.opt.adam method), 15update_stats() (combo.blm.core.model.model method),
19
Wwrite() (combo.search.discrete.policy.policy method), 14write() (combo.search.discrete.results.history method),
14
XX (combo.variable.variable attribute), 11
ZZ (combo.variable.variable attribute), 11
32 Index