Git Product home page Git Product logo

physbo's People

Contributors

bookcherry avatar k-kitai avatar k-yoshimi avatar kojitsuda avatar rtmr avatar tsuyos-u avatar yomichi avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar avatar avatar avatar avatar avatar avatar

Forkers

fermiq unixjunkie hikaruri nkato1206 bookcherry rossxli yuki-asano k-kitai yuenxq qup20 hengyuli-ozaki-lab chrinide keshavaspanda mizukihamada nkitamuraqc

physbo's Issues

initial_data指定時のsimulator実行失敗

PHYSBO V.1.0.1

physbo.search.discrete.policyでinitial_dataを指定した際、policy.random_searchでsimulatorを用いると下記エラーが発生します。
TypeError: only integer scalar arrays can be converted to a scalar index

通常、ndarray型のpolicy.actionsがlist型になっており、policy.actionsをndarrayにキャストするとエラーが回避できました。
ソースを確認したところ、policy.__init__が原因であると思われます。
ご確認のほどよろしくお願いいたします。

Code 
import physbo
import numpy as np

def load_data():
    A =  np.asarray(np.loadtxt('./example.csv',skiprows=1, delimiter=',') )
    X = A[:,0:3]
    t  = -A[:,3]
    return X, t

X, t = load_data()
X = physbo.misc.centering(X)

class simulator:
    def __init__(self, t):
        self.t = t
    
    def __call__(self, action):
        return -self.t[action]

import random
random.seed(0)
calculated_ids = random.sample(range(t.size), 20)
t_initial = t[calculated_ids]

policy = physbo.search.discrete.policy(test_X=X, initial_data=[calculated_ids, t_initial])
actions = policy.random_search(max_num_probes=10, simulator=simulator(t))

physbo.search.discrete.policy.bayes_searchのログ出力

is_dispをFalseにしても下記のようにログが出力されます。
ログを出力しないようにしたいのですが、何が原因でしょうか？

[実行環境]
windows10 64bit
Python 3.7.4
anaconda version 1.7.2
physbo == 1.0.0
Jupyter Notebook

policy.bayes_search(
    training=None,
    max_num_probes=50,
    num_search_each_probe=1,
    predictor=None,
    is_disp=False,
    simulator=None,
    score='EI',
    interval=0,
    num_rand_basis=0
)

Start the hyper parameter learning ...
0 -th epoch marginal likelihood 7.913191638026905
50 -th epoch marginal likelihood 7.913191500165695
100 -th epoch marginal likelihood 7.913191496779653
150 -th epoch marginal likelihood 7.913191496756556
200 -th epoch marginal likelihood 7.913191496756403
250 -th epoch marginal likelihood 7.913191496756403
300 -th epoch marginal likelihood 7.913191496756403
350 -th epoch marginal likelihood 7.913191496756404
400 -th epoch marginal likelihood 7.913191496756404
450 -th epoch marginal likelihood 7.913191496756404
500 -th epoch marginal likelihood 7.913191496756404
Done

Fail to load s5-210.csv

PHYSBOマニュアルにサンプルデータ取得先の記載がなさそうだったため、COMBOに記載のURLからs5-210.csvをダウンロードしました。

http://www.tsudalab.org/files/s5-210.csv

その後、PHYSBOマニュアルの「ガウス過程」のコードを実行したところ、サンプルデータに文字列が入っているのでエラーが発生しました。
サンプルデータの内訳の記載がマニュアルに見当たらないため、どのような変更が正しいのかわかりません。
ご教授いただけないでしょうか？

import physbo

import numpy as np

def load_data():
    A =  np.asarray(np.loadtxt('./s5-210.csv',skiprows=1, delimiter=',') )
    X = A[:,0:3]
    t  = -A[:,3]
    return X, t

X, t = load_data()
X = physbo.misc.centering( X )

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-20-fa9f3786555e> in <module>
      9     return X, t
     10 
---> 11 X, t = load_data()
     12 X = physbo.misc.centering( X )

<ipython-input-20-fa9f3786555e> in load_data()
      4 
      5 def load_data():
----> 6     A =  np.asarray(np.loadtxt('./s5-210.csv',skiprows=1, delimiter=',') )
      7     X = A[:,0:3]
      8     t  = -A[:,3]

~\AppData\Local\Continuum\anaconda3\lib\site-packages\numpy\lib\npyio.py in loadtxt(fname, dtype, comments, delimiter, converters, skiprows, usecols, unpack, ndmin, encoding, max_rows, like)
   1144         # converting the data
   1145         X = None
-> 1146         for x in read_data(_loadtxt_chunksize):
   1147             if X is None:
   1148                 X = np.array(x, dtype)

~\AppData\Local\Continuum\anaconda3\lib\site-packages\numpy\lib\npyio.py in read_data(chunk_size)
    995 
    996             # Convert each value according to its column and store
--> 997             items = [conv(val) for (conv, val) in zip(converters, vals)]
    998 
    999             # Then pack it according to the dtype's nesting

~\AppData\Local\Continuum\anaconda3\lib\site-packages\numpy\lib\npyio.py in <listcomp>(.0)
    995 
    996             # Convert each value according to its column and store
--> 997             items = [conv(val) for (conv, val) in zip(converters, vals)]
    998 
    999             # Then pack it according to the dtype's nesting

~\AppData\Local\Continuum\anaconda3\lib\site-packages\numpy\lib\npyio.py in floatconv(x)
    732         if '0x' in x:
    733             return float.fromhex(x)
--> 734         return float(x)
    735 
    736     typ = dtype.type

ValueError: could not convert string to float: '"(gb1'

`numpy==1.19`使用時の`import physbo`実行失敗

numpyのバージョン1.19を使用してphysboを使用したいのですが、import physboでnumpyが原因のValueErrorが出てしまいます。
numpy 1.19でも動かせるように対応していただくことは可能でしょうか？
また、回避策などご存じでしたら、ご教示いただけますと幸いです。

エラーの再現方法

  1. python環境の作成
  2. physbo, numpyのインストール
    pip3 install physbo==1.1.0 numpy==1.19.5
  3. pythonスクリプト、或いはnotebookでimport physboを実行

想定される挙動

以下、1行目にimport physboがあるpythonスクリプトを実行した際のエラーメッセージです。

Traceback (most recent call last):
  File "test.py", line 1, in <module>
    import physbo
  File "/home/yoshizawa/.pyenv/versions/miniconda3-4.2.12/envs/physbo_test/lib/python3.7/site-packages/physbo/__init__.py", line 1, in <module>
    from . import gp
  File "/home/yoshizawa/.pyenv/versions/miniconda3-4.2.12/envs/physbo_test/lib/python3.7/site-packages/physbo/gp/__init__.py", line 1, in <module>
    from . import cov
  File "/home/yoshizawa/.pyenv/versions/miniconda3-4.2.12/envs/physbo_test/lib/python3.7/site-packages/physbo/gp/cov/__init__.py", line 1, in <module>
    from .gauss import gauss
  File "/home/yoshizawa/.pyenv/versions/miniconda3-4.2.12/envs/physbo_test/lib/python3.7/site-packages/physbo/gp/cov/gauss.py", line 4, in <module>
    from ._src.enhance_gauss import grad_width64
  File "physbo/gp/cov/_src/enhance_gauss.pyx", line 1, in init physbo.gp.cov._src.enhance_gauss
ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject

実行環境

  • OS: CentOS 7
  • Python: 3.7.13
  • physbo: 1.1.0
  • numpy: 1.19.5
  • scipy: 1.7.3

よろしくお願いいたします。

AttributeError: module 'physbo.search.utility' has no attribute 'show_start_message_multi_search_mo'

PHYSBO v1.01

多目的最適化チュートリアルにおいて、num_search_each_probeに１より大きい値を設定すると以下のエラーが表示されます
utility.pyにはshow_start_message_multi_search関数は存在しますが、show_start_message_multi_search_mo関数は存在しないようです。
ご確認よろしくおねがいします

res_TS = policy.bayes_search(num_search_each_probe=2,max_num_probes=10, simulator=simu, score='TS', interval=10, num_rand_basis=100)

AttributeError: module 'physbo.search.utility' has no attribute 'show_start_message_multi_search_mo'

ファイルからロードした後に新しい結果を書き込めない

historyクラスの内容をsaveメソッドでnpzファイルに保存しておいたものをloadメソッドで読み込み、複数のアクションの結果を書き込もうとするとエラーになります。再現コードは以下のとおりです。

from physbo.search.discrete import history
his = history()
his.write(t=[3.14], action=[0])
his.save("history.npz")
his.load("history.npz")
his.write(t=[2.18, 1.618], action=[1, 2]) # ValueError: could not broadcast input array from shape (2,) into shape (0,)

原因はhistoryクラスのint型のインスタンス変数であるnum_runsとtotal_num_search が、npzファイルに保存した際にnumpy.ndarray型に変化してしまい、その値を別の変数に格納するときに参照渡しになっているからのようです。loadメソッドでこれらの値を読み出した時に、キャストで明示的にint型に戻してはいかがでしょうか。

saveしたpolicyをloadした時、同じactionが呼ばれてしまう

saveして保存したpolicyをloadし直した時、一度呼ばれて評価値を登録したactionが再度呼ばれてしまうような気がするのですが、いかがでしょうか？
評価値が未登録のactionが優先的に呼ばれる仕様が望ましく感じるのですがいかがでしょうか？

以下、自作プログラムをprintした挙動になるのですが、

chosen_actions
 [12  6  7  9  8 10 18  1 19 13  4 15 14 11  2 16  3  5 17]

のpolicyをloadした場合において、random_searchを実行すると、
(この時は)[11]のchosen_actionに入っているactionが再度呼ばれる結果となっております。

Start the initial hyper parameter searching ...
Done

Start the hyper parameter learning ...
0 -th epoch marginal likelihood -34.930600301307116
50 -th epoch marginal likelihood -39.134970116321696
100 -th epoch marginal likelihood -41.25847547249403
150 -th epoch marginal likelihood -42.341124562570016
200 -th epoch marginal likelihood -42.93564581049239
250 -th epoch marginal likelihood -43.29619395568853
300 -th epoch marginal likelihood -43.541479385193696
350 -th epoch marginal likelihood -43.727110572165785
400 -th epoch marginal likelihood -43.878704166640375
450 -th epoch marginal likelihood -44.00773411736684
500 -th epoch marginal likelihood -44.11926492568572
Done

chosen_actions:
 [12  6  7  9  8 10 18  1 19 13  4 15 14 11  2 16  3  5 17]

Guess next param by RANDOM search
 Next index: [11]
 Next param: [10 10 5 200 20 20 3000 450 30 30]

Fail to import physbo using anaconda navigator

For windows OS, the problem failure to import physbo using anaconda navigator has been reported.
When you see the following error, please uninstall numpy and install numpy again (numpy>=1.20 works).
After that, import physbo again.

File "physbo\gp\cov_src\enhance_gauss.pyx", line 1, in init physbo.gp.cov._src.enhance_gauss
ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject

獲得関数値の出力

星健夫(鳥取大)です．
2020年度ISSP高度化プロジェクト(ソフト名：2DMAT)を通じて，PHYSBOを使わせていただいています．ベイズ最適化の途中において，（PHYSBO内部で生成される）獲得関数（全メッシュ上の関数値）の出力ができると良いです．何iterationごとにdumpする，と言うような設定を想定しています．

manual でのparameter selection

現状、parameter searchの方法として、

の２通りが実装されているかと思います。
これらに加え、ユーザがmanualでactionを指定できる機能が実装されると有用に感じたのですがいかがでしょうか？
今後実装の予定はございますでしょうか？

実験からのinteractive実行でPHYSBOを活用させていただいているのですが、登録しそこねたり失敗した実験を再実験したい場合に、actionのmanual selection機能があると有用かと感じました。

physbo.search.discrete_multi.policyのインタラクティブ実行時の挙動

多目的最適化の計算をインタラクティブ実行すると、以下のような挙動になりました。

  1. physbo.search.utility.show_search_results_moの履歴に最初の一個目の結果が表示されない（表示数が2以上の場合）
  2. policy.get_post_fmean実行時にValueError: shapes (100,10) and (9,) not aligned: 10 (dim 1) != 9 (dim 0)のようなエラーが発生

なお、2.のエラーに関してはsimulatorクラスを定義して実行した場合には発生しませんでした。
設定の間違いや回避策などがあればご教示いただければ幸いです。

実行環境

  • Windows10 64bit
  • Python 3.10.7
  • PHYSBO 1.1.0
  • JupyterLab

テストコード（インタラクティブ実行）

import numpy as np
import physbo

def f(x):
    y1 = 1 + np.exp(-x)
    y2 = 2*x**2 + 1
    return np.c_[-y1, -y2]

x = np.linspace(0.1,5,100)
test_X = x.reshape(len(x),1)

policy = physbo.search.discrete_multi.policy(test_X=test_X, num_objectives=2)
policy.set_seed(0)

# ランダムサーチ
num_rand = 5
for i in range(num_rand):
    actions = policy.random_search(max_num_probes=1, simulator=None)
    t = f(test_X[actions[0]])
    policy.write(actions, t)
    physbo.search.utility.show_search_results_mo(policy.history, 10)

# ベイズ最適化
num_bo = 5
for i in range(num_bo):
    actions = policy.bayes_search(max_num_probes=1, simulator=None, score='HVPI', interval=0)
    t = f(test_X[actions[0]])
    policy.write(actions, t)
    physbo.search.utility.show_search_results_mo(policy.history, 10)

post_fmean = policy.get_post_fmean(test_X) # ここでエラー発生
post_fcov = policy.get_post_fcov(test_X)

エラーメッセージ全文

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In [1], line 31
     28     policy.write(actions, t)
     29     physbo.search.utility.show_search_results_mo(policy.history, 10)
---> 31 post_fmean = policy.get_post_fmean(test_X)
     32 post_fcov = policy.get_post_fcov(test_X)

File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\physbo\search\discrete_multi\policy.py:293, in policy.get_post_fmean(self, xs)
    291         predictor_list[i].fit(self.training_list[i], 0)
    292         predictor_list[i].prepare(self.training_list[i])
--> 293 fmean = [
    294     predictor.get_post_fmean(training, X)
    295     for predictor, training in zip(predictor_list, self.training_list)
    296 ]
    297 return np.array(fmean).T

File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\physbo\search\discrete_multi\policy.py:294, in <listcomp>(.0)
    291         predictor_list[i].fit(self.training_list[i], 0)
    292         predictor_list[i].prepare(self.training_list[i])
    293 fmean = [
--> 294     predictor.get_post_fmean(training, X)
    295     for predictor, training in zip(predictor_list, self.training_list)
    296 ]
    297 return np.array(fmean).T

File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\physbo\gp\predictor.py:98, in predictor.get_post_fmean(self, training, test)
     96 if self.model.stats is None:
     97     self.prepare(training)
---> 98 return self.model.get_post_fmean(training.X, test.X)

File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\physbo\gp\core\model.py:248, in model.get_post_fmean(self, X, Z, params)
    245     params = np.copy(self.params)
    247 if self.inf == "exact":
--> 248     post_fmu = inf.exact.get_post_fmean(self, X, Z, params)
    250 return post_fmu

File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\physbo\gp\inf\exact.py:182, in get_post_fmean(gp, X, Z, params)
    179 fmu = gp.prior.get_mean(ntest)
    180 G = gp.prior.get_cov(X=Z, Z=X, params=prior_params)
--> 182 return G.dot(alpha) + fmu

ValueError: shapes (100,10) and (9,) not aligned: 10 (dim 1) != 9 (dim 0)

simulatorクラスを使った場合

del policy

class simulator(object):
    def __init__(self, X):
        self.t = f(X)

    def __call__( self, action):
        return self.t[action]

simu = simulator(test_X)

policy = physbo.search.discrete_multi.policy(test_X=test_X, num_objectives=2)
policy.set_seed(0)
policy.random_search(max_num_probes=5, simulator=simu)
res = policy.bayes_search(max_num_probes=5, simulator=simu, score='HVPI', interval=0)

post_fmean = policy.get_post_fmean(test_X)
post_fcov = policy.get_post_fcov(test_X)

post_fmean[0:res.num_runs] # こちらは問題なく実行可能

Fail to install physbo

physboでPEP517エラーがでてしまい、インストールに失敗します。

【実行環境】
OS: Windows10 64bit
python 3.7.4
Microsoft(R) C/C++ Optimizing Compiler Version 19.28.29336 for x86

C:\Users\user\Desktop>pip install physbo
Collecting physbo
  Using cached physbo-1.0.0.tar.gz (34 kB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
    Preparing wheel metadata ... done
Requirement already satisfied: numpy in c:\users\user\appdata\local\continuum\anaconda3\lib\site-packages (from physbo) (1.20.2)
Requirement already satisfied: scipy in c:\users\user\appdata\local\continuum\anaconda3\lib\site-packages (from physbo) (1.6.2)
Building wheels for collected packages: physbo
  Building wheel for physbo (PEP 517) ... error
  ERROR: Command errored out with exit status 1:
   command: 'c:\users\user\appdata\local\continuum\anaconda3\python.exe' 'c:\users\user\appdata\local\continuum\anaconda3\lib\site-packages\pip\_vendor\pep517\_in_process.py' build_wheel 'C:\Users\ab012462\AppData\Local\Temp\tmpfebzpugi'
       cwd: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0
  Complete output (105 lines):
  running bdist_wheel
  running build
  running build_py
  creating build
  creating build\lib.win-amd64-3.7
  creating build\lib.win-amd64-3.7\physbo
  copying physbo\predictor.py -> build\lib.win-amd64-3.7\physbo
  copying physbo\variable.py -> build\lib.win-amd64-3.7\physbo
  copying physbo\__init__.py -> build\lib.win-amd64-3.7\physbo
  creating build\lib.win-amd64-3.7\physbo\blm
  copying physbo\blm\predictor.py -> build\lib.win-amd64-3.7\physbo\blm
  copying physbo\blm\__init__.py -> build\lib.win-amd64-3.7\physbo\blm
  creating build\lib.win-amd64-3.7\physbo\gp
  copying physbo\gp\predictor.py -> build\lib.win-amd64-3.7\physbo\gp
  copying physbo\gp\__init__.py -> build\lib.win-amd64-3.7\physbo\gp
  creating build\lib.win-amd64-3.7\physbo\misc
  copying physbo\misc\centering.py -> build\lib.win-amd64-3.7\physbo\misc
  copying physbo\misc\gauss_elim.py -> build\lib.win-amd64-3.7\physbo\misc
  copying physbo\misc\set_config.py -> build\lib.win-amd64-3.7\physbo\misc
  copying physbo\misc\__init__.py -> build\lib.win-amd64-3.7\physbo\misc
  creating build\lib.win-amd64-3.7\physbo\opt
  copying physbo\opt\adam.py -> build\lib.win-amd64-3.7\physbo\opt
  copying physbo\opt\__init__.py -> build\lib.win-amd64-3.7\physbo\opt
  creating build\lib.win-amd64-3.7\physbo\search
  copying physbo\search\pareto.py -> build\lib.win-amd64-3.7\physbo\search
  copying physbo\search\score.py -> build\lib.win-amd64-3.7\physbo\search
  copying physbo\search\score_multi.py -> build\lib.win-amd64-3.7\physbo\search
  copying physbo\search\utility.py -> build\lib.win-amd64-3.7\physbo\search
  copying physbo\search\__init__.py -> build\lib.win-amd64-3.7\physbo\search
  creating build\lib.win-amd64-3.7\physbo\blm\basis
  copying physbo\blm\basis\fourier.py -> build\lib.win-amd64-3.7\physbo\blm\basis
  copying physbo\blm\basis\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\basis
  creating build\lib.win-amd64-3.7\physbo\blm\core
  copying physbo\blm\core\model.py -> build\lib.win-amd64-3.7\physbo\blm\core
  copying physbo\blm\core\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\core
  creating build\lib.win-amd64-3.7\physbo\blm\inf
  copying physbo\blm\inf\exact.py -> build\lib.win-amd64-3.7\physbo\blm\inf
  copying physbo\blm\inf\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\inf
  creating build\lib.win-amd64-3.7\physbo\blm\lik
  copying physbo\blm\lik\gauss.py -> build\lib.win-amd64-3.7\physbo\blm\lik
  copying physbo\blm\lik\linear.py -> build\lib.win-amd64-3.7\physbo\blm\lik
  copying physbo\blm\lik\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\lik
  creating build\lib.win-amd64-3.7\physbo\blm\prior
  copying physbo\blm\prior\gauss.py -> build\lib.win-amd64-3.7\physbo\blm\prior
  copying physbo\blm\prior\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\prior
  creating build\lib.win-amd64-3.7\physbo\blm\lik\_src
  copying physbo\blm\lik\_src\cov.py -> build\lib.win-amd64-3.7\physbo\blm\lik\_src
  copying physbo\blm\lik\_src\__init__.py -> build\lib.win-amd64-3.7\physbo\blm\lik\_src
  creating build\lib.win-amd64-3.7\physbo\gp\core
  copying physbo\gp\core\learning.py -> build\lib.win-amd64-3.7\physbo\gp\core
  copying physbo\gp\core\model.py -> build\lib.win-amd64-3.7\physbo\gp\core
  copying physbo\gp\core\prior.py -> build\lib.win-amd64-3.7\physbo\gp\core
  copying physbo\gp\core\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\core
  creating build\lib.win-amd64-3.7\physbo\gp\cov
  copying physbo\gp\cov\gauss.py -> build\lib.win-amd64-3.7\physbo\gp\cov
  copying physbo\gp\cov\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\cov
  creating build\lib.win-amd64-3.7\physbo\gp\inf
  copying physbo\gp\inf\exact.py -> build\lib.win-amd64-3.7\physbo\gp\inf
  copying physbo\gp\inf\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\inf
  creating build\lib.win-amd64-3.7\physbo\gp\lik
  copying physbo\gp\lik\gauss.py -> build\lib.win-amd64-3.7\physbo\gp\lik
  copying physbo\gp\lik\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\lik
  creating build\lib.win-amd64-3.7\physbo\gp\mean
  copying physbo\gp\mean\const.py -> build\lib.win-amd64-3.7\physbo\gp\mean
  copying physbo\gp\mean\zero.py -> build\lib.win-amd64-3.7\physbo\gp\mean
  copying physbo\gp\mean\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\mean
  creating build\lib.win-amd64-3.7\physbo\gp\cov\_src
  copying physbo\gp\cov\_src\__init__.py -> build\lib.win-amd64-3.7\physbo\gp\cov\_src
  creating build\lib.win-amd64-3.7\physbo\misc\_src
  copying physbo\misc\_src\__init__.py -> build\lib.win-amd64-3.7\physbo\misc\_src
  creating build\lib.win-amd64-3.7\physbo\search\discrete
  copying physbo\search\discrete\policy.py -> build\lib.win-amd64-3.7\physbo\search\discrete
  copying physbo\search\discrete\results.py -> build\lib.win-amd64-3.7\physbo\search\discrete
  copying physbo\search\discrete\__init__.py -> build\lib.win-amd64-3.7\physbo\search\discrete
  creating build\lib.win-amd64-3.7\physbo\search\discrete_multi
  copying physbo\search\discrete_multi\policy.py -> build\lib.win-amd64-3.7\physbo\search\discrete_multi
  copying physbo\search\discrete_multi\results.py -> build\lib.win-amd64-3.7\physbo\search\discrete_multi
  copying physbo\search\discrete_multi\__init__.py -> build\lib.win-amd64-3.7\physbo\search\discrete_multi
  running build_ext
  cythoning physbo/misc/_src/traceAB.pyx to physbo/misc/_src\traceAB.c
  cythoning physbo/misc/_src/cholupdate.pyx to physbo/misc/_src\cholupdate.c
  cythoning physbo/misc/_src/diagAB.pyx to physbo/misc/_src\diagAB.c
  cythoning physbo/gp/cov/_src/enhance_gauss.pyx to physbo/gp/cov/_src\enhance_gauss.c
  cythoning physbo/misc/_src/logsumexp.pyx to physbo/misc/_src\logsumexp.c
  building 'physbo.misc._src.traceAB' extension
  creating build\temp.win-amd64-3.7
  creating build\temp.win-amd64-3.7\Release
  creating build\temp.win-amd64-3.7\Release\physbo
  creating build\temp.win-amd64-3.7\Release\physbo\misc
  creating build\temp.win-amd64-3.7\Release\physbo\misc\_src
  C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29333\bin\HostX86\x64\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD -IC:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\numpy\core\include -Ic:\users\user\appdata\local\continuum\anaconda3\include -Ic:\users\user\appdata\local\continuum\anaconda3\include "-IC:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29333\include" /Tcphysbo/misc/_src\traceAB.c /Fobuild\temp.win-amd64-3.7\Release\physbo/misc/_src\traceAB.obj -O3
  cl : コマンド ライン warning D9002 : 不明なオプション '-O3' を無視します。
  traceAB.c
  c:\users\user\appdata\local\continuum\anaconda3\include\pyconfig.h(59): fatal error C1083: include ファイルを開けません。'io.h':No such file or directory
  C:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\Cython\Compiler\Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0\physbo\misc\_src\traceAB.pyx
    tree = Parsing.p_module(s, pxd, full_module_name)
  C:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\Cython\Compiler\Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0\physbo\misc\_src\cholupdate.pyx
    tree = Parsing.p_module(s, pxd, full_module_name)
  C:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\Cython\Compiler\Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0\physbo\misc\_src\diagAB.pyx
    tree = Parsing.p_module(s, pxd, full_module_name)
  C:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\Cython\Compiler\Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0\physbo\gp\cov\_src\enhance_gauss.pyx
    tree = Parsing.p_module(s, pxd, full_module_name)
  C:\Users\user\AppData\Local\Temp\pip-build-env-wd9fvi1n\overlay\Lib\site-packages\Cython\Compiler\Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: C:\Users\user\AppData\Local\Temp\pip-install-9yf6cee5\physbo_9c0b7923470a4d1b91fe8f3c800440f0\physbo\misc\_src\logsumexp.pyx
    tree = Parsing.p_module(s, pxd, full_module_name)
  error: command 'C:\\Program Files (x86)\\Microsoft Visual Studio\\2019\\BuildTools\\VC\\Tools\\MSVC\\14.28.29333\\bin\\HostX86\\x64\\cl.exe' failed with exit status 2
  ----------------------------------------
  ERROR: Failed building wheel for physbo
Failed to build physbo
ERROR: Could not build wheels for physbo which use PEP 517 and cannot be installed directly

How to adjust the parameters of the optimization process?

Hi, thank you very much for your work, this library helped me a lot. I found that I am not getting the desired optimization result during the optimization process, I input the verified optimal result into the search range but I can't optimize to get this optimal result. I guess the parameter settings of the optimization process need to be adjusted, but I don't know how to do it and which parameters I can adjust? Can you help me with this problem? Thank you!

獲得関数TSでのget_scoreの戻り値

獲得関数がEIやPIのときは戻り値が一重のリストですが、TSのときは二重のリストで戻ってきます。どちらかに統一してもらえないでしょうか？

EI, PIの場合
[2.90841329e-204 ... 3.72840278e-189]

TSの場合
[[-3.21928677 ... -4.84607939]]

多次元のプロット関数

多次元でのガウス過程回帰を用いた実験計画法を実施したいのですが、フィッティングしたガウス過程回帰と獲得関数を２次元サーフェスプロットで表示する機能はないでしょうか？

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.