y-mitsui

/* ベイズ情報量基準 */
double bic(double logLike,int numSample,int numModel){
 return -2*logLike+numModel*log(numSample);
}
/* 学習スコア */
double score(int numNode,int *edge,int numSample,double **likelies,int *variablePattern){
 double like;
 int i,j,numModel,numParents;
 unsigned char parentPattern;

y-mitsui

"""
G検定による独立性の検定
"""
import numpy as np
from scipy.stats import chisqprob, chisquare

def gtest(f_obs, f_exp=None, ddof=0):
    """
    http://en.wikipedia.org/wiki/G-test
    The G test can test for goodness of fit to a distribution

y-mitsui

# Refererce:http://ushi-goroshi.hatenablog.com/entry/2018/01/11/233317
# install.packages("bnlearn")
# install.packages("BNSL")

library(bnlearn)
library(BNSL)
set.seed(123)


norm <- rnorm(4000)

y-mitsui

import numpy as np

n_sample = 1000
true_coef = np.array([2.0, 3.0])
u0 = np.array([0.0, 0.0])
A0 = 0.1 * np.eye(2)
sample_X = np.random.normal(size=(n_sample, 2))
sample_y = np.dot(sample_X, true_coef.reshape(-1, 1)).reshape(-1,) + np.random.normal(size=n_sample)

tmp1 = np.linalg.inv(np.dot(sample_X.T, sample_X))

y-mitsui

"""
String Kernel
Reference: "５．構造化データに対するカーネル" 福水健次
http://www.ism.ac.jp/~fukumizu/ISM_lecture_2010/Kernel_5_structured.pdf
"""
import numpy as np

def subSequenceJustLength(result, str, cur_str, length, depth=0):
    if depth == length:
        result.append(cur_str)

y-mitsui

# -*- coding: utf-8 -*-
"""
Reference:
    トピックモデルによる統計的潜在意味解析 4章
"""

from __future__ import print_function
from scipy.special import digamma
import numpy as np
import time

y-mitsui

# -*- coding: utf-8 -*-
"""
Reference:
    トピックモデルによる統計的潜在意味解析 4章
"""

from __future__ import print_function
from scipy.special import digamma
import numpy as np
import time

y-mitsui

'''
Created on 2017/05/24

'''

from sklearn.grid_search import GridSearchCV
import numpy as np
from sklearn.cross_validation import KFold
from sklearn.metrics import mean_squared_error

y-mitsui

import pystan
import numpy as np

stan_code = """
data {
    int<lower=0> N;
    int<lower=0> K;
    matrix[N,K] x;
    vector[N] y;
}

y-mitsui

import pystan
import numpy as np

stan_code = """
data {
    int<lower=0> N;
    real<lower=0> x[N];
}
parameters {
    real mu;