oiehot

//
// ES6 Features
//
// 원본: http://es6-features.org
//

// 1. 상수
// =======

const PI = 3.141593

oiehot

import numpy as np

'''
class ReluLayer:
    def __init__(self):
        self.x = None
    def forward(self, x):
        self.x = x
        if x > 0:
            return x

oiehot

class AddLayer:
    def __init__(self):
        pass
    def forward(self, x, y):
        return x + y
    def backward(self, dout):
        dx = dout * 1
        dy = dout * 1
        return dx, dy

oiehot

import sys, os
sys.path.append(os.pardir)
import numpy as np
from common.functions import *
from common.gradient import *
from dataset.mnist import load_mnist

class TwoLayerNet:
    def __init__(self, inputSize, hiddenSize, outputSize, weight=0.01):
        self.w1 = weight * np.random.randn(inputSize, hiddenSize)

oiehot

# 머신러닝, 수치 미분, 편미분, 기울기gradient, 경사 하강법gradient descent

# 수치 미분numerical differentiation, 전방 차분
def numerical_diff_bad(f, x):
    h = 10e-50 # np.float32(1e-50) => 0.0, 반올림 오차 문제.
    return (f(x+h) - f(x)) / h

# 수치 미분, 중심 차분
def numerical_diff(f, x):
    h = 1e-4 # 0.0001

oiehot

global proc TZ_Transform_CopyPivot(string $src, string $dest) {
	float $scalePivot[3];
	float $rotatePivot[3];

	$scalePivot  = `xform -query -worldSpace -scalePivot $src`;
	$rotatePivot = `xform -query -worldSpace -rotatePivot $src`;
	
	xform -worldSpace -scalePivot  $scalePivot[0] $scalePivot[1] $scalePivot[2]    $dest;
	xform -worldSpace -rotatePivot $rotatePivot[0] $rotatePivot[1] $rotatePivot[2] $dest;
}

oiehot

// 0: Smooth
// 1: Jelly
// 2: Round Rect 1
// 3: Round Rect 2
// 4: Bevel
// 5: Hard
global proc TZ_Subdiv_SetLevel(int $level)
{
    int $minSubdivLevel = max(5, $level); // 최저 Subdiv레벨: 5

oiehot

import sys, os
import numpy as np

# 정답표, One hot label
t = np.array([
	[0,0,1,0,0,0,0,0,0,0], # 2
	[0,0,0,1,0,0,0,0,0,0], # 3
	[0,0,0,0,0,1,0,0,0,0], # 5
	[1,0,0,0,0,0,0,0,0,0], # 0
	[0,1,0,0,0,0,0,0,0,0]  # 1

oiehot

import sys, os
import numpy as np
import pickle
import random
from PIL import Image

# https://github.com/oreilly-japan/deep-learning-from-scratch/common, dataset
sys.path.append(os.pardir)
from dataset.mnist import load_mnist
from common.functions import sigmoid, softmax

oiehot

import numpy as np

def step(x):
    return np.array( x > 0, dtype=np.int )

def sigmoid(x):
    y = 1 / (1 + np.exp(-x))
    return y

def relu(x):