python

Exception

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

def error01():
    a=10
    a/0
    #ZeroDivisionError: division by zero

def error02():
    a= [1, 2, 3, 4, 5]
    a[10]
    #IndexError: list index out of range

def error03():
    a = 1000
    a + "Hello"
    #TypeError: unsupported operand type(s) for +: 'int' and 'str'

def error04():
    a=10
    a+b
    #NameError: name 'b' is not defined

if __name__ == "__main__":
    error01()
    error02()
    error03()
    error04()
    print("program is done")

• 크롤링 코드를 작성 했을때

  “https://sports.news.naver.com/news?oid=109&aid=0004526080" : 페이지있음,
  “https://sports.news.naver.com/news?oid=109&aid=0004526081" : 페이지가 없다면,
  “https://sports.news.naver.com/news?oid=109&aid=0005526080" : 페이지 있음,

• 크롤링 코드 멈춤
• 프로그램이 멈춰서 안됨

Exeption의 종류

java 의 try catch 구문과 같음

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

def try_func(x, idx):
    try:
        return 100/x[idx]
    except ZeroDivisionError:
        print("did't divide zero")
    except IndexError:
        print("not in range of Index")
    except TypeError:
        print("there is type Error")
    except NameError:
        print("it is not definated parameter")
    finally:
        print("무조건 실행됨")


def main():
    a = [50, 60, 0, 70]
    print(try_func(a,1))

    # Zero Division Error
    print(try_func(a,0))

    # Index Error
    print(try_func(a,5))

    # type Error
    print(try_func(a, "hi"))


if __name__ == "__main__":
    main()

어떻게던 프로그램이 돌아 갈 수 있도록
만들어 주는 것이 중요하다.

• class 정리

1. __init__ : set_name, set_id 해 주지 않고, 통합시켜주는 역할
2. __eq__, __ne__ : 부등호 연산자
3. 상속, 다형성(서로다른 클래스에서 공통으로 쓰는 함수)
4. Exception
5. class attribute / instance attribute / instance method 차이
6. 추상 class (안배웠음)
7. data incapsulation

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

class SalaryExcept(ValueError): pass # 상속
class TipExept(SalaryExcept): pass # 상속

class Employee:

    MIN_SALARY = 30000
    MAX_Bonus = 20000

    def __init__(self, name, salary = 30000):
        self.name = name
        if salary< Employee.MIN_SALARY:
            raise SalaryExcept("급여가 너무 낮아요!")
        self.salary = salary

    def give_bonus(self, amount):
        if amount > Employee.MAX_Bonus:
            print("보너스가 너무 많아 ")
        elif self.salary + amount < Employee.MIN_SALARY :
            print("보너스 지급 후의 급여도 매우 낮다. ")
        else:
            self.salary += amount

if __name__ == "__main__":
    emp = Employee("YH", salary= 10000)

    try:
        emp.give_bonus(70000)
    except SalaryExcept:
        print("Error Salary")

    try:
        emp.give_bonus(-10000)
    except tipExcept:
        print("Error Tip")

여전히 Error가 나는 코드
나는 Exception 안됨

python

Bank _ 계좌 만들기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

class Human:

    def __init__(self, name):
        self.name = name


if __name__ == "__main__":
    human01 = Human(name="A")
    human02 = Human(name="A")

    print(human01 == human02)
    print("human 01 : ", human01)
    print("human 02 : ", human02)

False
human 01 : <__main__.Human object at 0x000001686E41CC10>
human 02 : <__main__.Human object at 0x000001686E41CE50>

저장되는 장소가 다르기 때문에 다르다.

Bank _ customer ID 확인하기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

class Bank:

    #instance attribute
    def __init__(self, cust_id, balance=0):
        self.balance = balance
        self.cust_id = cust_id

    #instance methode
    def withdraw(self, amount):
        self.balance -= amount

    def __eq__(self, other):
        print("__eq()__ is called")
        return self.cust_id == other.cust_id

if __name__ == "__main__":
    account01 = Bank(123, 1000)
    account02 = Bank(123, 1000)
    account03 = Bank(456, 1000)
    print(account01 == account02)
    print(account02 == account03)
    print(account01 == account03)

eq() is called
True
eq() is called
False
eq() is called
False

• 부등호 연산자
  • != : ne()
  • >= : ge()
  • <= : le()
  • > : gt()
  • < : lt()

eq() 함수 사용하기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8

class Bank:

    #instance attribute
    def __init__(self, cust_id, balance=0):
        self.balance, self.cust_id = balance, cust_id


    #instance methode
    def withdraw(self, amount):
        self.balance -= amount

    def __eq__(self, other):
        print("__eq()__ is called")
        return (self.cust_id == other.cust_id) and (type(self) == type(other))

class Phone:

    def __init__(self, cust_id):
        self.cust_id = cust_id

    def __eq__(self, other):
        return self.cust_id == other.cust_id


if __name__ == "__main__":
    account01 = Bank(1234)
    phone01 = Phone(1234)

    print(account01 == phone01)

eq() is called
False

eq를 불러와서 같은지 확인 할 수 있다.

접근기록, log 기록 확인하기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8


class Bank:
    def __init__(self, cust_id, name, balance = 0):
        self.cust_id, self.name, self.balance = cust_id, name, balance

    def __str__(self):
        cust_str = """
        customer:
            cust_id : {cust_id}
            name : {name}
            balance : {balance}
        """.format(cust_id = self.cust_id, name = self.name, balance= self.balance)

        return cust_str

if __name__ == "__main__":
    bank_cust = Bank(123, "YH")
    print(bank_cust)

• DB에 저장 되지 않지만, 로그 기록을 확인 할 수 있다.

str() and repr() 비교

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

# /c/Users/brill/Desktop/PyThon_Function/venv/Scripts/python
# -*- coding : UTF-8


class Bank:
    def __init__(self, cust_id, name, balance = 0):
        self.cust_id, self.name, self.balance = cust_id, name, balance

    def __str__(self):
        cust_str = """
        customer:
            cust_id : {cust_id}
            name : {name}
            balance : {balance}
        """.format(cust_id = self.cust_id, name = self.name, balance= self.balance)

        return cust_str

    def __repr__(self):
        cust_str = "Bank({cust_id}, '{name}', {balance})".format(cust_id = self.cust_id, name = self.name, balance= self.balance)
        return cust_str

if __name__ == "__main__":
    bank_cust = Bank(123, "YH")
    print(str(bank_cust))
    print(repr(bank_cust))

difference of str() and repr()

machine Learning Model Algoridms

• 비 선형 모델 : KNN,
• 선형 모델 :

Decision Tree MachineLearning

Introduction

• 과적합 : 모델의 정확도만 높이기 위해 분류 조건(depth)만 강조하여 실제 상황에서 유연하게 대처하는 능력이 떨어지게 되는 문제가 발생하게 되는것.
• 가지치기(pruning)을 통해 유연성을 유지.
  • Max_depth를 대략적으로 잡아서 (3, 5, 10…) RMS 값 비교
  • Random search
  • 하이퍼파라미터 (grid Search)

• 분류기준 (수식은 아래서 책에서 확인)

  1. 정보이득 :
     • 자식노드의 불순도가 낮을 수록 정보의 이득이 커진다.(효율성 Up)
     • 정보 이득이 높은 속성을 기준으로 알아서 나누어 준다.
     1. 엔트로피의 정의 :
        • 엔트로피는 높을 수록 좋다.
     2. 지니불순도 :
        • 순도는 높을 수록 좋다.
     3. 분류오차 :
        • 어떤 시나리오가 더 좋은가에 대한 계산
        • 1이 되면 균등, 완벽하게 나누어 졌다고

• ㅇㅇ

공식은 이쪽에 가면 있다.

계산은 컴퓨터가 다 해준다.

우리는 보고 좋은 분류 기준을 선택 하며 됩니다.

분류기준 1. 분류 오차

분류기준 2. 지니 불순도

분류기준 2. 엔트로피

• 정보이득을 최대로 하는 옵션을 찾는다.

실습

1
2
3
4
5
6
7
8

from sklearn import datasets 
import numpy as np 

iris = datasets.load_iris()
X = iris.data[:, [2, 3]]
y = iris.target 

print("클래스 레이블:", np.unique(y))

클래스 레이블: [0 1 2]

1
2
3
4
5
6
7

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size = 0.3, random_state = 1
)

print("y 레이블 갯수:", np.bincount(y))

y 레이블 갯수: [50 50 50]

시각화

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

from matplotlib.colors import ListedColormap
import matplotlib.pyplot as plt

def plot_decision_regions(X, y, classifier, test_idx=None, resolution=0.02):

    # 마커와 컬러맵을 설정합니다.
    markers = ('s', 'x', 'o', '^', 'v')
    colors = ('red', 'blue', 'lightgreen', 'gray', 'cyan')
    cmap = ListedColormap(colors[:len(np.unique(y))])

    # 결정 경계를 그립니다.
    x1_min, x1_max = X[:, 0].min() - 1, X[:, 0].max() + 1
    x2_min, x2_max = X[:, 1].min() - 1, X[:, 1].max() + 1
    xx1, xx2 = np.meshgrid(np.arange(x1_min, x1_max, resolution),
                           np.arange(x2_min, x2_max, resolution))
    Z = classifier.predict(np.array([xx1.ravel(), xx2.ravel()]).T)
    Z = Z.reshape(xx1.shape)
    plt.contourf(xx1, xx2, Z, alpha=0.3, cmap=cmap)
    plt.xlim(xx1.min(), xx1.max())
    plt.ylim(xx2.min(), xx2.max())

    for idx, cl in enumerate(np.unique(y)):
        plt.scatter(x=X[y == cl, 0], 
                    y=X[y == cl, 1],
                    alpha=0.8, 
                    c=colors[idx],
                    marker=markers[idx], 
                    label=cl, 
                    edgecolor='black')

    # 테스트 샘플을 부각하여 그립니다.
    if test_idx:
        X_test, y_test = X[test_idx, :], y[test_idx]

        plt.scatter(X_test[:, 0],
                    X_test[:, 1],
                    c='',
                    edgecolor='black',
                    alpha=1.0,
                    linewidth=1,
                    marker='o',
                    s=100, 
                    label='test set')

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

import matplotlib.pyplot as plt
import numpy as np

# 지니 불순도 함수
def gini(p):
    return p * (1 - p) + (1 - p) * (1 - (1 - p))


# 엔트로피 함수 
def entropy(p):
    return - p * np.log2(p) - (1 - p) * np.log2((1 - p))

# 분류 오차
def error(p):
    return 1 - np.max([p, 1 - p])

x = np.arange(0.0, 1.0, 0.01)

ent = [entropy(p) if p != 0 else None for p in x]
sc_ent = [e * 0.5 if e else None for e in ent]
err = [error(i) for i in x]

fig = plt.figure()
ax = plt.subplot(111)
for i, lab, ls, c, in zip([ent, sc_ent, gini(x), err], 
                          ['Entropy', 'Entropy (scaled)', 
                           'Gini Impurity', 'Misclassification Error'],
                          ['-', '-', '--', '-.'],
                          ['black', 'lightgray', 'red', 'green', 'cyan']):
    line = ax.plot(x, i, label=lab, linestyle=ls, lw=2, color=c)

ax.legend(loc='upper center', bbox_to_anchor=(0.5, 1.15),
          ncol=5, fancybox=True, shadow=False)

ax.axhline(y=0.5, linewidth=1, color='k', linestyle='--')
ax.axhline(y=1.0, linewidth=1, color='k', linestyle='--')
plt.ylim([0, 1.1])
plt.xlabel('p(i=1)')
plt.ylabel('Impurity Index')
plt.show()

• 정보 이득을 최대로 하는 옵션을 찾아서

1
2
3
4

from sklearn.tree import DecisionTreeClassifier

tree_gini = DecisionTreeClassifier(criterion="gini", max_depth=3)
tree_gini.fit(X_train, y_train)

DecisionTreeClassifier(max_depth=3)

• depth를 3으로 해 주었기 때문에 과적합 X

1
2
3
4
5
6
7
8
9

X_combined = np.vstack((X_train, X_test))
y_combined = np.hstack((y_train, y_test))

plot_decision_regions(X_combined, y_combined, classifier=tree_gini, test_idx = range(105, 150))
plt.xlabel("petal length")
plt.ylabel("petal width")
plt.legend(loc = "upper left")
plt.tight_layout()
plt.show()

1
2
3
4
5
6
7
8
9
10
11
12
13
14

from pydotplus import graph_from_dot_data
from sklearn.tree import export_graphviz

dot_data = export_graphviz(tree_gini,
                           filled=True, 
                           rounded=True,
                           class_names=['Setosa', 
                                        'Versicolor',
                                        'Virginica'],
                           feature_names=['petal length', 
                                          'petal width'],
                           out_file=None) 
graph = graph_from_dot_data(dot_data) 
graph.write_png('gini_tree.png') 

True

• gini 로 1개 Entripy 로 1개 짜서 해야함

• gini: default
• Entropy : 도 해보고 비교

1
2
3
4
5
6
7
8
9
10
11
12

tree_entropy = DecisionTreeClassifier(criterion="entropy", max_depth=3)
tree_entropy.fit(X_train, y_train)

X_combined = np.vstack((X_train, X_test))
y_combined = np.hstack((y_train, y_test))

plot_decision_regions(X_combined, y_combined, classifier=tree_entropy, test_idx = range(105, 150))
plt.xlabel("petal length")
plt.ylabel("petal width")
plt.legend(loc = "upper left")
plt.tight_layout()
plt.show()

• 모형을 도식화로

1
2
3
4
5
6
7
8
9
10
11
12
13
14

from pydotplus import graph_from_dot_data
from sklearn.tree import export_graphviz

dot_data = export_graphviz(tree_entropy,
                           filled=True, 
                           rounded=True,
                           class_names=['Setosa', 
                                        'Versicolor',
                                        'Virginica'],
                           feature_names=['petal length', 
                                          'petal width'],
                           out_file=None) 
graph = graph_from_dot_data(dot_data) 
graph.write_png('Entropy_tree.png') 

entropy가 0이 되면 더이상 나눌 필요가 없다.

• sklearn에서는 분류오차는 없다.

• 지니 와 엔트로피 두개를 보고 더 나은 것을 선택

머신러닝 배우기

<아직 안배운 부분>

• 스태킹 알고리즘 (앙상블)

AutoMachineLearning by pycaret

pycaret

pycaret으로 autoML 하기

• low-code machine learning library
• PyCaret 2.0 ver.
  • 분석가가 가야 하는 최종 도착지
  • 머신러닝 + operation (운영) : 배포 ->
    • MLflow, Airflow, Kubeflow…

gitHub and pycaret

pycaret install

1
2
3

!pip install pycaret

# !pip install pycaret==2.0

Collecting pycaret
Downloading pycaret-2.3.5-py3-none-any.whl (288 kB)
|████████████████████████████████| 288 kB 5.4 MB/s
Collecting lightgbm>=2.3.1
Downloading lightgbm-3.3.1-py3-none-manylinux1_x86_64.whl (2.0 MB)
|████████████████████████████████| 2.0 MB 54.5 MB/s
Collecting pyod
Downloading pyod-0.9.5.tar.gz (113 kB)
|████████████████████████████████| 113 kB 67.4 MB/s
Requirement already satisfied: textblob in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.15.3)
Requirement already satisfied: yellowbrick>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.3.post1)
Collecting Boruta
Downloading Boruta-0.3-py3-none-any.whl (56 kB)
|████████████████████████████████| 56 kB 4.6 MB/s
Collecting pyLDAvis
Downloading pyLDAvis-3.3.1.tar.gz (1.7 MB)
|████████████████████████████████| 1.7 MB 63.6 MB/s
Installing build dependencies … done
Getting requirements to build wheel … done
Installing backend dependencies … done
Preparing wheel metadata … done
Requirement already satisfied: seaborn in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.11.2)
Requirement already satisfied: nltk in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.2.5)
Collecting imbalanced-learn==0.7.0
Downloading imbalanced_learn-0.7.0-py3-none-any.whl (167 kB)
|████████████████████████████████| 167 kB 65.5 MB/s
Requirement already satisfied: numpy==1.19.5 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.19.5)
Collecting kmodes>=0.10.1
Downloading kmodes-0.11.1-py2.py3-none-any.whl (19 kB)
Requirement already satisfied: spacy<2.4.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (2.2.4)
Collecting umap-learn
Downloading umap-learn-0.5.2.tar.gz (86 kB)
|████████████████████████████████| 86 kB 4.8 MB/s
Requirement already satisfied: IPython in /usr/local/lib/python3.7/dist-packages (from pycaret) (5.5.0)
Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.2.2)
Requirement already satisfied: wordcloud in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.5.0)
Requirement already satisfied: cufflinks>=0.17.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.17.3)
Collecting mlflow
Downloading mlflow-1.22.0-py3-none-any.whl (15.5 MB)
|████████████████████████████████| 15.5 MB 68.5 MB/s
Collecting mlxtend>=0.17.0
Downloading mlxtend-0.19.0-py2.py3-none-any.whl (1.3 MB)
|████████████████████████████████| 1.3 MB 66.9 MB/s
Requirement already satisfied: pandas in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.1.5)
Collecting scikit-plot
Downloading scikit_plot-0.3.7-py3-none-any.whl (33 kB)
Requirement already satisfied: joblib in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.1.0)
Requirement already satisfied: plotly>=4.4.1 in /usr/local/lib/python3.7/dist-packages (from pycaret) (4.4.1)
Collecting scikit-learn==0.23.2
Downloading scikit_learn-0.23.2-cp37-cp37m-manylinux1_x86_64.whl (6.8 MB)
|████████████████████████████████| 6.8 MB 37.1 MB/s
Requirement already satisfied: gensim<4.0.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.6.0)
Collecting pandas-profiling>=2.8.0
Downloading pandas_profiling-3.1.0-py2.py3-none-any.whl (261 kB)
|████████████████████████████████| 261 kB 60.5 MB/s
Requirement already satisfied: ipywidgets in /usr/local/lib/python3.7/dist-packages (from pycaret) (7.6.5)
Requirement already satisfied: scipy<=1.5.4 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.4.1)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn==0.23.2->pycaret) (3.0.0)
Requirement already satisfied: setuptools>=34.4.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (57.4.0)
Requirement already satisfied: colorlover>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (0.3.0)
Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (1.15.0)
Requirement already satisfied: smart-open>=1.2.1 in /usr/local/lib/python3.7/dist-packages (from gensim<4.0.0->pycaret) (5.2.1)
Requirement already satisfied: pickleshare in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (0.7.5)
Requirement already satisfied: pexpect in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (4.8.0)
Requirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (5.1.1)
Requirement already satisfied: pygments in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (2.6.1)
Requirement already satisfied: simplegeneric>0.8 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (0.8.1)
Requirement already satisfied: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (1.0.18)
Requirement already satisfied: decorator in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (4.4.2)
Requirement already satisfied: nbformat>=4.2.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (5.1.3)
Requirement already satisfied: widgetsnbextension=3.5.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (3.5.2)
Requirement already satisfied: ipython-genutils=0.2.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (0.2.0)
Requirement already satisfied: ipykernel>=4.5.1 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (4.10.1)
Requirement already satisfied: jupyterlab-widgets>=1.0.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (1.0.2)
Requirement already satisfied: jupyter-client in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets->pycaret) (5.3.5)
Requirement already satisfied: tornado>=4.0 in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets->pycaret) (5.1.1)
Requirement already satisfied: wheel in /usr/local/lib/python3.7/dist-packages (from lightgbm>=2.3.1->pycaret) (0.37.0)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (3.0.6)
Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (2.8.2)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (1.3.2)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (0.11.0)
Requirement already satisfied: jsonschema!=2.5.0,>=2.4 in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets->pycaret) (2.6.0)
Requirement already satisfied: jupyter-core in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets->pycaret) (4.9.1)
Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas->pycaret) (2018.9)
Collecting visions[type_image_path]==0.7.4
Downloading visions-0.7.4-py3-none-any.whl (102 kB)
|████████████████████████████████| 102 kB 12.4 MB/s
Collecting pydantic>=1.8.1
Downloading pydantic-1.8.2-cp37-cp37m-manylinux2014_x86_64.whl (10.1 MB)
|████████████████████████████████| 10.1 MB 24.8 MB/s
Collecting tangled-up-in-unicode==0.1.0
Downloading tangled_up_in_unicode-0.1.0-py3-none-any.whl (3.1 MB)
|████████████████████████████████| 3.1 MB 22.1 MB/s
Collecting joblib
Downloading joblib-1.0.1-py3-none-any.whl (303 kB)
|████████████████████████████████| 303 kB 60.4 MB/s
Collecting requests>=2.24.0
Downloading requests-2.26.0-py2.py3-none-any.whl (62 kB)
|████████████████████████████████| 62 kB 805 kB/s
Requirement already satisfied: tqdm>=4.48.2 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (4.62.3)
Collecting PyYAML>=5.0.0
Downloading PyYAML-6.0-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl (596 kB)
|████████████████████████████████| 596 kB 42.5 MB/s
Requirement already satisfied: missingno>=0.4.2 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (0.5.0)
Requirement already satisfied: markupsafe=2.0.1 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (2.0.1)
Collecting htmlmin>=0.1.12
Downloading htmlmin-0.1.12.tar.gz (19 kB)
Collecting multimethod>=1.4
Downloading multimethod-1.6-py3-none-any.whl (9.4 kB)
Collecting phik>=0.11.1
Downloading phik-0.12.0-cp37-cp37m-manylinux2010_x86_64.whl (675 kB)
|████████████████████████████████| 675 kB 41.5 MB/s
Requirement already satisfied: jinja2>=2.11.1 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (2.11.3)
Requirement already satisfied: networkx>=2.4 in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (2.6.3)
Requirement already satisfied: attrs>=19.3.0 in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (21.2.0)
Requirement already satisfied: Pillow in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (7.1.2)
Collecting imagehash
Downloading ImageHash-4.2.1.tar.gz (812 kB)
|████████████████████████████████| 812 kB 37.7 MB/s
Collecting scipy<=1.5.4
Downloading scipy-1.5.4-cp37-cp37m-manylinux1_x86_64.whl (25.9 MB)
|████████████████████████████████| 25.9 MB 74.1 MB/s
Requirement already satisfied: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly>=4.4.1->pycaret) (1.3.3)
Requirement already satisfied: wcwidth in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->IPython->pycaret) (0.2.5)
Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.7/dist-packages (from pydantic>=1.8.1->pandas-profiling>=2.8.0->pycaret) (3.10.0.2)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (1.24.3)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2021.10.8)
Requirement already satisfied: charset-normalizer=2.0.0 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2.0.8)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2.10)
Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (3.0.6)
Requirement already satisfied: srsly<1.1.0,>=1.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.5)
Requirement already satisfied: wasabi<1.1.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (0.8.2)
Requirement already satisfied: thinc==7.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (7.4.0)
Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.6)
Requirement already satisfied: plac<1.2.0,>=0.9.6 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.1.3)
Requirement already satisfied: catalogue<1.1.0,>=0.0.7 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.0)
Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (2.0.6)
Requirement already satisfied: blis<0.5.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (0.4.1)
Requirement already satisfied: importlib-metadata>=0.20 in /usr/local/lib/python3.7/dist-packages (from catalogue<1.1.0,>=0.0.7->spacy<2.4.0->pycaret) (4.8.2)
Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=0.20->catalogue<1.1.0,>=0.0.7->spacy<2.4.0->pycaret) (3.6.0)
Requirement already satisfied: notebook>=4.4.1 in /usr/local/lib/python3.7/dist-packages (from widgetsnbextension=3.5.0->ipywidgets->pycaret) (5.3.1)
Requirement already satisfied: terminado>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.12.1)
Requirement already satisfied: nbconvert in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (5.6.1)
Requirement already satisfied: Send2Trash in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (1.8.0)
Requirement already satisfied: pyzmq>=13 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel>=4.5.1->ipywidgets->pycaret) (22.3.0)
Requirement already satisfied: ptyprocess in /usr/local/lib/python3.7/dist-packages (from terminado>=0.8.1->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.7.0)
Requirement already satisfied: PyWavelets in /usr/local/lib/python3.7/dist-packages (from imagehash->visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (1.2.0)
Collecting querystring-parser
Downloading querystring_parser-1.2.4-py2.py3-none-any.whl (7.9 kB)
Requirement already satisfied: entrypoints in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (0.3)
Collecting alembic<=1.4.1
Downloading alembic-1.4.1.tar.gz (1.1 MB)
|████████████████████████████████| 1.1 MB 66.5 MB/s
Collecting gitpython>=2.1.0
Downloading GitPython-3.1.24-py3-none-any.whl (180 kB)
|████████████████████████████████| 180 kB 40.6 MB/s
Requirement already satisfied: protobuf>=3.7.0 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (3.17.3)
Requirement already satisfied: sqlalchemy in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.4.27)
Requirement already satisfied: Flask in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.1.4)
Requirement already satisfied: cloudpickle in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.3.0)
Collecting databricks-cli>=0.8.7
Downloading databricks-cli-0.16.2.tar.gz (58 kB)
|████████████████████████████████| 58 kB 5.6 MB/s
Collecting prometheus-flask-exporter
Downloading prometheus_flask_exporter-0.18.6-py3-none-any.whl (17 kB)
Requirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (21.3)
Requirement already satisfied: click>=7.0 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (7.1.2)
Requirement already satisfied: sqlparse>=0.3.1 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (0.4.2)
Collecting gunicorn
Downloading gunicorn-20.1.0-py3-none-any.whl (79 kB)
|████████████████████████████████| 79 kB 7.6 MB/s
Collecting docker>=4.0.0
Downloading docker-5.0.3-py2.py3-none-any.whl (146 kB)
|████████████████████████████████| 146 kB 58.9 MB/s
Collecting Mako
Downloading Mako-1.1.6-py2.py3-none-any.whl (75 kB)
|████████████████████████████████| 75 kB 4.2 MB/s
Collecting python-editor>=0.3
Downloading python_editor-1.0.4-py3-none-any.whl (4.9 kB)
Requirement already satisfied: tabulate>=0.7.7 in /usr/local/lib/python3.7/dist-packages (from databricks-cli>=0.8.7->mlflow->pycaret) (0.8.9)
Collecting websocket-client>=0.32.0
Downloading websocket_client-1.2.3-py3-none-any.whl (53 kB)
|████████████████████████████████| 53 kB 1.2 MB/s
Collecting gitdb<5,>=4.0.1
Downloading gitdb-4.0.9-py3-none-any.whl (63 kB)
|████████████████████████████████| 63 kB 1.6 MB/s
Collecting smmap<6,>=3.0.1
Downloading smmap-5.0.0-py3-none-any.whl (24 kB)
Requirement already satisfied: greenlet!=0.4.17 in /usr/local/lib/python3.7/dist-packages (from sqlalchemy->mlflow->pycaret) (1.1.2)
Requirement already satisfied: itsdangerous<2.0,>=0.24 in /usr/local/lib/python3.7/dist-packages (from Flask->mlflow->pycaret) (1.1.0)
Requirement already satisfied: Werkzeug<2.0,>=0.15 in /usr/local/lib/python3.7/dist-packages (from Flask->mlflow->pycaret) (1.0.1)
Requirement already satisfied: bleach in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (4.1.0)
Requirement already satisfied: pandocfilters>=1.4.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (1.5.0)
Requirement already satisfied: testpath in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.5.0)
Requirement already satisfied: mistune<2,>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.8.4)
Requirement already satisfied: defusedxml in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.7.1)
Requirement already satisfied: webencodings in /usr/local/lib/python3.7/dist-packages (from bleach->nbconvert->notebook>=4.4.1->widgetsnbextension=3.5.0->ipywidgets->pycaret) (0.5.1)
Requirement already satisfied: prometheus-client in /usr/local/lib/python3.7/dist-packages (from prometheus-flask-exporter->mlflow->pycaret) (0.12.0)
Collecting pyLDAvis
Downloading pyLDAvis-3.3.0.tar.gz (1.7 MB)
|████████████████████████████████| 1.7 MB 44.1 MB/s
Installing build dependencies … done
Getting requirements to build wheel … done
Installing backend dependencies … done
Preparing wheel metadata … done
Downloading pyLDAvis-3.2.2.tar.gz (1.7 MB)
|████████████████████████████████| 1.7 MB 30.5 MB/s
Requirement already satisfied: numexpr in /usr/local/lib/python3.7/dist-packages (from pyLDAvis->pycaret) (2.7.3)
Requirement already satisfied: future in /usr/local/lib/python3.7/dist-packages (from pyLDAvis->pycaret) (0.16.0)
Collecting funcy
Downloading funcy-1.16-py2.py3-none-any.whl (32 kB)
Requirement already satisfied: numba>=0.35 in /usr/local/lib/python3.7/dist-packages (from pyod->pycaret) (0.51.2)
Requirement already satisfied: statsmodels in /usr/local/lib/python3.7/dist-packages (from pyod->pycaret) (0.10.2)
Requirement already satisfied: llvmlite<0.35,>=0.34.0.dev0 in /usr/local/lib/python3.7/dist-packages (from numba>=0.35->pyod->pycaret) (0.34.0)
Requirement already satisfied: patsy>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from statsmodels->pyod->pycaret) (0.5.2)
Collecting pynndescent>=0.5
Downloading pynndescent-0.5.5.tar.gz (1.1 MB)
|████████████████████████████████| 1.1 MB 55.1 MB/s
Building wheels for collected packages: htmlmin, imagehash, alembic, databricks-cli, pyLDAvis, pyod, umap-learn, pynndescent
Building wheel for htmlmin (setup.py) … done
Created wheel for htmlmin: filename=htmlmin-0.1.12-py3-none-any.whl size=27098 sha256=6dff1694390dae41ea8bd3ca00f5564142023ea037fa606be0a8ffba9c16d1da
Stored in directory: /root/.cache/pip/wheels/70/e1/52/5b14d250ba868768823940c3229e9950d201a26d0bd3ee8655
Building wheel for imagehash (setup.py) … done
Created wheel for imagehash: filename=ImageHash-4.2.1-py2.py3-none-any.whl size=295207 sha256=9e38b104e77871b6f6a6a9267c3debd3ac85d39441acb3cda64d4dc07a11dd27
Stored in directory: /root/.cache/pip/wheels/4c/d5/59/5e3e297533ddb09407769762985d134135064c6831e29a914e
Building wheel for alembic (setup.py) … done
Created wheel for alembic: filename=alembic-1.4.1-py2.py3-none-any.whl size=158172 sha256=652d8b88b2468cf1d1c9f1c3242dda689e0f670bc6e5b88dc4dbf087fecbaccc
Stored in directory: /root/.cache/pip/wheels/be/5d/0a/9e13f53f4f5dfb67cd8d245bb7cdffe12f135846f491a283e3
Building wheel for databricks-cli (setup.py) … done
Created wheel for databricks-cli: filename=databricks_cli-0.16.2-py3-none-any.whl size=106811 sha256=9dbaaca3ece5f6a1522d676d8b1ec35c065a0bd0c564bd862ae3012984b70c9a
Stored in directory: /root/.cache/pip/wheels/f4/5c/ed/e1ce20a53095f63b27b4964abbad03e59cf3472822addf7d29
Building wheel for pyLDAvis (setup.py) … done
Created wheel for pyLDAvis: filename=pyLDAvis-3.2.2-py2.py3-none-any.whl size=135618 sha256=471d50a9a2e725465ffc7d32b21edb15c9b84dc0573891185a43e97f567aa0a7
Stored in directory: /root/.cache/pip/wheels/f8/b1/9b/560ac1931796b7303f7b517b949d2d31a4fbc512aad3b9f284
Building wheel for pyod (setup.py) … done
Created wheel for pyod: filename=pyod-0.9.5-py3-none-any.whl size=132699 sha256=d116f5b46155bf0fa31aa88cc21da0e3be461b448e9c9b2d599c763a5ef0a6a1
Stored in directory: /root/.cache/pip/wheels/3d/bb/b7/62b60fb451b33b0df1ab8006697fba7a6a49709a629055cf77
Building wheel for umap-learn (setup.py) … done
Created wheel for umap-learn: filename=umap_learn-0.5.2-py3-none-any.whl size=82709 sha256=7c48e34d2c19d333a623ed12491d3c7d07bafd52f2d35e474df56908f5cc7525
Stored in directory: /root/.cache/pip/wheels/84/1b/c6/aaf68a748122632967cef4dffef68224eb16798b6793257d82
Building wheel for pynndescent (setup.py) … done
Created wheel for pynndescent: filename=pynndescent-0.5.5-py3-none-any.whl size=52603 sha256=7abff97eebc36deea7220f1b5e9907020826a07404003a9c7d794fef4d396e87
Stored in directory: /root/.cache/pip/wheels/af/e9/33/04db1436df0757c42fda8ea6796d7a8586e23c85fac355f476
Successfully built htmlmin imagehash alembic databricks-cli pyLDAvis pyod umap-learn pynndescent
Installing collected packages: tangled-up-in-unicode, smmap, scipy, multimethod, joblib, websocket-client, visions, scikit-learn, requests, python-editor, Mako, imagehash, gitdb, querystring-parser, PyYAML, pynndescent, pydantic, prometheus-flask-exporter, phik, htmlmin, gunicorn, gitpython, funcy, docker, databricks-cli, alembic, umap-learn, scikit-plot, pyod, pyLDAvis, pandas-profiling, mlxtend, mlflow, lightgbm, kmodes, imbalanced-learn, Boruta, pycaret
Attempting uninstall: scipy
Found existing installation: scipy 1.4.1
Uninstalling scipy-1.4.1:
Successfully uninstalled scipy-1.4.1
Attempting uninstall: joblib
Found existing installation: joblib 1.1.0
Uninstalling joblib-1.1.0:
Successfully uninstalled joblib-1.1.0
Attempting uninstall: scikit-learn
Found existing installation: scikit-learn 1.0.1
Uninstalling scikit-learn-1.0.1:
Successfully uninstalled scikit-learn-1.0.1
Attempting uninstall: requests
Found existing installation: requests 2.23.0
Uninstalling requests-2.23.0:
Successfully uninstalled requests-2.23.0
Attempting uninstall: PyYAML
Found existing installation: PyYAML 3.13
Uninstalling PyYAML-3.13:
Successfully uninstalled PyYAML-3.13
Attempting uninstall: pandas-profiling
Found existing installation: pandas-profiling 1.4.1
Uninstalling pandas-profiling-1.4.1:
Successfully uninstalled pandas-profiling-1.4.1
Attempting uninstall: mlxtend
Found existing installation: mlxtend 0.14.0
Uninstalling mlxtend-0.14.0:
Successfully uninstalled mlxtend-0.14.0
Attempting uninstall: lightgbm
Found existing installation: lightgbm 2.2.3
Uninstalling lightgbm-2.2.3:
Successfully uninstalled lightgbm-2.2.3
Attempting uninstall: imbalanced-learn
Found existing installation: imbalanced-learn 0.8.1
Uninstalling imbalanced-learn-0.8.1:
Successfully uninstalled imbalanced-learn-0.8.1
ERROR: pip’s dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
google-colab 1.0.0 requires requests=2.23.0, but you have requests 2.26.0 which is incompatible.
datascience 0.10.6 requires folium==0.2.1, but you have folium 0.8.3 which is incompatible.
albumentations 0.1.12 requires imgaug<0.2.7,>=0.2.5, but you have imgaug 0.2.9 which is incompatible.
Successfully installed Boruta-0.3 Mako-1.1.6 PyYAML-6.0 alembic-1.4.1 databricks-cli-0.16.2 docker-5.0.3 funcy-1.16 gitdb-4.0.9 gitpython-3.1.24 gunicorn-20.1.0 htmlmin-0.1.12 imagehash-4.2.1 imbalanced-learn-0.7.0 joblib-1.0.1 kmodes-0.11.1 lightgbm-3.3.1 mlflow-1.22.0 mlxtend-0.19.0 multimethod-1.6 pandas-profiling-3.1.0 phik-0.12.0 prometheus-flask-exporter-0.18.6 pyLDAvis-3.2.2 pycaret-2.3.5 pydantic-1.8.2 pynndescent-0.5.5 pyod-0.9.5 python-editor-1.0.4 querystring-parser-1.2.4 requests-2.26.0 scikit-learn-0.23.2 scikit-plot-0.3.7 scipy-1.5.4 smmap-5.0.0 tangled-up-in-unicode-0.1.0 umap-learn-0.5.2 visions-0.7.4 websocket-client-1.2.3

pycaret을 그냥 설치 할 수도 있고,
version을 정해서 설치 할 수도 있다.

• 먼저 개괄적으로 확인만 하고, github 에서 autoML을 pycaret 2.0 ver로 진행 해야 한다.

google colab에 설치 한 경우

Install 후 런타임>런타임다시시작(CTRL+M) 을 꼭 한번 해 준 후

아래 설명에 따라 가는 것이 좋음.

** 만약 오류가 난다면, 런타임 초기화 후 import, 런타임다시시작 후 진행 하는 것을 추천

• 왜 그런지 모름 _ 그냥 이렇게 하면 된다는 것만 알려주겠음

Data Load

1
2

from pycaret.datasets import get_data
data = get_data("diamond")

pycaret.regression

1
2
3

from pycaret.regression import *
reg_set = setup(data, target = 'Price', transform_target = True, 
                log_experiment = True, experiment_name = 'diamond')

• pycaret.regression : 종류

모델 만들기

• 최적의 모델을 만들기 위해 한줄의 코드면 된다 ㅠㅠ

1

best  = compare_models()

모형 평가

• 최적의 모델 확인 후 평가 역시 코드 한줄 ㅠㅠ 감동

1

plot_model(best)

1

plot_model(best, plot = "feature")

모형 저장, 모형 배포

• MLOps 개념, RestAPI, Flask

1
2
3
4

finalize_best = finalize_model(best)

#save model
save_model(finalize_best, "diamond_pipeline")

Transformation Pipeline and Model Successfully Saved
(Pipeline(memory=None,
steps=[(‘dtypes’,
DataTypes_Auto_infer(categorical_features=[],
display_types=True, features_todrop=[],
id_columns=[], ml_usecase=’regression’,
numerical_features=[], target=’Price’,
time_features=[])),
(‘imputer’,
Simple_Imputer(categorical_strategy=’not_available’,
fill_value_categorical=None,
fill_value_numerical=None,
numeric_strategy=’…
learning_rate=0.1,
max_depth=-1,
min_child_samples=20,
min_child_weight=0.001,
min_split_gain=0.0,
n_estimators=100,
n_jobs=-1,
num_leaves=31,
objective=None,
random_state=2882,
reg_alpha=0.0,
reg_lambda=0.0,
silent=’warn’,
subsample=1.0,
subsample_for_bin=200000,
subsample_freq=0),
silent=’warn’, subsample=1.0,
subsample_for_bin=200000,
subsample_freq=0)]],
verbose=False), ‘diamond_pipeline.pkl’)

MLOps

• devOPs (개발과 운영 팀이 별도로 있었음.)
• 자동화 되면서 같이 됨.
  MLOps dash board

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

!pip install mlflow --quiet
!pip install pyngrok --quiet

import mlflow

with mlflow.start_run(run_name="MLflow on Colab"):
  mlflow.log_metric("m1", 2.0)
  mlflow.log_param("p1", "mlflow-colab")

# run tracking UI in the background
get_ipython().system_raw("mlflow ui --port 5000 &") # run tracking UI in the background


# create remote tunnel using ngrok.com to allow local port access
# borrowed from https://colab.research.google.com/github/alfozan/MLflow-GBRT-demo/blob/master/MLflow-GBRT-demo.ipynb#scrollTo=4h3bKHMYUIG6

from pyngrok import ngrok

# Terminate open tunnels if exist
ngrok.kill()

# Setting the authtoken (optional)
# Get your authtoken from https://dashboard.ngrok.com/auth
NGROK_AUTH_TOKEN = ""
ngrok.set_auth_token(NGROK_AUTH_TOKEN)

# Open an HTTPs tunnel on port 5000 for http://localhost:5000
ngrok_tunnel = ngrok.connect(addr="5000", proto="http", bind_tls=True)
print("MLflow Tracking UI:", ngrok_tunnel.public_url)

배포 하는 것이 마지막 단계인데, 구글 코랩에서 안먹는 다는 것이 함정이라고 한다.

언젠간 내가 스스로 할 수 있는 날이 오지 않을까 한다.

Chepter 2 _사이킷런으로 시작하는 머신러닝(04)

파이썬 머신러닝 완벽 가이드
ref. & copyright(c) Book

Model Selection 모듈 소개

• 사이킷런 (scikit-Learn) : 파이썬 머신러닝 라이브러리 중 가장 많이 사용되는 라이브러리

1
2
3
4
5

import sklearn

print(cklearn.__version__)

Kaggle 주택가격 예측

Kaggle house oruces advabced regression

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

# This Python 3 environment comes with many helpful analytics libraries installed
# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python
# For example, here's several helpful packages to load

import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)

# Input data files are available in the read-only "../input/" directory
# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory

import os
for dirname, _, filenames in os.walk('/kaggle/input'):
    for filename in filenames:
        print(os.path.join(dirname, filename))

# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using "Save & Run All" 
# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session

데이터 다운로드 및 불러오기

1
2
3
4
5
6
7

import pandas as pd

train = pd.read_csv("/kaggle/input/house-prices-advanced-regression-techniques/train.csv")
test = pd.read_csv("/kaggle/input/house-prices-advanced-regression-techniques/test.csv")

train.shape, test.shape
#변수를 줄여야 겠다. 어떤 변수를 줄여야 할까 ?

EDA

• 이상치과 중복값 제거
• overallQual (주택의 상태를 1~10등급으로 책정)
• 평점 1 : 판매가가 높음 = 이상치라고 판단 할 수 있따. : 이걸 제거 해 줘야 함.

1
2

train.info()
#80개 컬럼, SalePrice(독립변수)를 제거 하고는 나머지가 종속변수 = 너무 많다 !! 

1
2
3

train.drop(train[(train['OverallQual'] < 5) & (train['SalePrice']> 200000)].index, inplace = True)
train.reset_index(drop = True, inplace = True)
train.shape

종속변수 시각화

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

import seaborn as sns
import matplotlib.pyplot as plt
from scipy.stats import norm

(mu, sigma) = norm.fit(train['SalePrice'])
print("The value of mu before log transformation is:", mu)
print("The value of sigma before log transformation is:", sigma)

fig, ax = plt.subplots(figsize=(10, 6))
sns.histplot(train['SalePrice'], color="b", stat="probability")
ax.xaxis.grid(False)
ax.set(ylabel="Frequency")
ax.set(xlabel="SalePrice")
ax.set(title="SalePrice distribution")

plt.axvline(mu, color='r', linestyle='--')
plt.text(mu + 10000, 0.11, 'Mean of SalePrice', rotation=0, color='r')
fig.show()

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

import numpy as np 

train["SalePrice"] = np.log1p(train["SalePrice"]) # 로그 변환 후 종속 변수 시각화 

(mu, sigma) = norm.fit(train['SalePrice'])
print("The value of mu before log transformation is:", mu)
print("The value of sigma before log transformation is:", sigma)

fig, ax = plt.subplots(figsize=(10, 6))
sns.histplot(train['SalePrice'], color="b", stat="probability")
ax.xaxis.grid(False)
ax.set(ylabel="Frequency")
ax.set(xlabel="SalePrice")
ax.set(title="SalePrice distribution")

plt.axvline(mu, color='r', linestyle='--')
plt.text(mu + 0.05, 0.111, 'Mean of SalePrice', rotation=0, color='r')
plt.ylim(0, 0.12)
fig.show()

data feature 제거

1. 모형 학습 시간 감소
2. 연산 시 noise 감소

그래서 : train ID 를 빼주기로함

1
2
3
4
5

train_ID = train['Id']
test_ID = test['Id']
train.drop(['Id'], axis=1, inplace=True)
test.drop(['Id'], axis=1, inplace=True)
train.shape, test.shape

1
2
3
4
5
6
7
8

# y 값 추출, dataset 분리할때 사용
y = train['SalePrice'].reset_index(drop=True)

# 뽑고 나면 원래 df에서 제거
train = train.drop('SalePrice', axis = 1)
train.shape, test.shape, y.shape

1
2
3
4
5

# data 합치기 
# - train data 와  Test를 같이 전처리 하기 위해 

all_df = pd.concat([train, test]).reset_index(drop=True)
all_df.shape

결측치 확인

• 결측치 처리
  1. 제거하기 : column 제거, 특정 행만 제거하기
  2. 채우기 : 1) numeric(수치형) : 평균 또는 중간값으로 채우기

         2) String(문자형) : 최빈값으로 채우기 
     

  3. 통계 기법이용, 채우기 (data 보간)
     • 실무에서는 (KNNImput)등, 시계열 자료 or 산업군에 따라 다르므로 가서 배워라.

1
2
3
4
5
6
7
8
9
10

#결측치 확인 

def check_na(data, head_num = 6):
  isnull_na = (data.isnull().sum() / len(data)) * 100
  data_na = isnull_na.drop(isnull_na[isnull_na == 0].index).sort_values(ascending=False)
  missing_data = pd.DataFrame({'Missing Ratio' :data_na, 
                               'Data Type': data.dtypes[data_na.index]})
  print("결측치 데이터 컬럼과 건수:\n", missing_data.head(head_num))

check_na(all_df, 20)

1
2
3
4

all_df.drop(['PoolQC', 'MiscFeature', 'Alley', 'Fence', 'FireplaceQu', 'LotFrontage'], axis=1, inplace=True)
check_na(all_df)

#아직도 결측치가 많이 있다. 

채우기

1. 문자열 채우기 
2. 

1. object column 추출

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

#a = all_df['BsmtCond'].value_counts().mode() #mode() : 최빈값 가장 빈도수가 높은 값 찾기
#a

print(all_df['BsmtCond'].value_counts())
print()
print(all_df['BsmtCond'].mode()[0])

#object column, 갯수 확인
import numpy as np
cat_all_vars = train.select_dtypes(exclude=[np.number]) #숫자인 것을 제외한 type column 이름 추출
print("The whole number of all_vars(문자형data)", len(list(cat_all_vars)))

#column 이름 뽑아내기 
final_cat_vars = []
for v in cat_all_vars:
    if v not in ['PoolQC', 'MiscFeature', 'Alley', 'Fence', 'FireplaceQu', 'LotFrontage']:
        final_cat_vars.append(v)

print("The whole number of final_cat_vars", len(final_cat_vars))

#최빈값을 찾아 넣어주기
for i in final_cat_vars:
  all_df[i] = all_df[i].fillna(all_df[i].mode()[0])

check_na(all_df, 20)
print("숫자형 data set의 결측치만 남은 것을 알 수 있다. ")

1
2
3
4
5
6
7
8
9
10
11
12

import numpy as np
num_all_vars = list(train.select_dtypes(include=[np.number]))
print("The whole number of all_vars", len(num_all_vars))

num_all_vars.remove('LotFrontage')

print("The whole number of final_cat_vars", len(num_all_vars))
for i in num_all_vars:
  all_df[i].fillna(value=all_df[i].median(), inplace=True)

print("결측치가 존재 하지 않은 것을 알 수 있다. ")
check_na(all_df, 20)

1

all_df.info()

왜도(Skewnewss) 처리하기 : 정규 분포를 이룰 수 있게 (설문조사 논문 통계의 경우 -1< 외도 <1)

• boxcose를 사용 할 예정

• 왜도가 양수일때, 음수일때 (좌, 우로 치우친 정도)

• 첨도가 양수일때, 음수일때 (뽀족한 정도)

• RMSE를 최적(낮게)으로 만들기 위해 조정.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

from scipy.stats import skew

#외도 판정을 받을 만한 data set을 확인 
def find_skew(x):
  return skew(x)
#앞에서 뽑은 numeric columns : num_all_vars
#사용자 정의함수를 쓰기 위해 apply(find_skew)를 사용, 오름차순정렬

skewness_features = all_df[num_all_vars].apply(find_skew).sort_values(ascending=False)
skewness_features

#high_skew = skew_valrs[slew_var > 1]

#0~1사이에 있는 것이 기준. 기준 밖으로 나간 경우 조정이 필요(정규분포를 만들어 주기 위해)
# 1. 박스코스 변환 : ML -> RMSE (2.5) 
# 2. 로그변환     : ML -> RMSE (2.1)
# => RMSE는 적은 것이 좋기 때문에, 로그 변환으로 사용 하는 것이 좋다. 

1
2
3
4
5
6
7
8
9
10
11
12
13
14

skewnewss_index = list(skewness_features.index)
skewnewss_index.remove('LotArea')
#외도 정도가 너무 높은 LotArea를 날려주는 것.
all_numeric_df = all_df.loc[:, skewnewss_index]


fig, ax = plt.subplots(figsize=(10, 6))
ax.set_xlim(0, all_numeric_df.max().sort_values(ascending=False)[0])
ax = sns.boxplot(data=all_numeric_df[skewnewss_index] , orient="h", palette="Set1")
ax.xaxis.grid(False)
ax.set(ylabel="Feature names")
ax.set(xlabel="Numeric values")
ax.set(title="Numeric Distribution of Features Before Box-Cox Transformation")
sns.despine(trim=True, left=True)

1
2
3
4
5
6
7
8
9
10
11
12

from scipy.special import boxcox1p
from scipy.stats import boxcox_normmax

high_skew = skewness_features[skewness_features > 1]
high_skew_index = high_skew.index

print("The data before Box-Cox Transformation: \n", all_df[high_skew_index].head())

for num_var in high_skew_index:
  all_df[num_var] = boxcox1p(all_df[num_var], boxcox_normmax(all_df[num_var] + 1))

print("The data after Box-Cox Transformation: \n", all_df[high_skew_index].head())

1
2
3
4
5
6
7
8

fig, ax = plt.subplots(figsize=(10, 6))
ax.set_xscale('log')
ax = sns.boxplot(data=all_df[high_skew_index] , orient="h", palette="Set1")
ax.xaxis.grid(False)
ax.set(ylabel="Feature names")
ax.set(xlabel="Numeric values")
ax.set(title="Numeric Distribution of Features Before Box-Cox Transformation")
sns.despine(trim=True, left=True)

도출 변수

Feature Engineering 의 Key step

• 판매량, 단가, 매출액 X
• 판매량 X 단가 = 매출액(New Value) : 도출 변수
  • ML은 수식이기 때문에 도출변수가 생성 되는것은 연산의 증가로 이어진다.
  • 시간이 오래 걸린다.
  • 결론 : 변수를 줄이는 것이 좋다.

1
2
3
4

#집의 층수 를 더해서 전체면적이라는 변수를 도출 
all_df['TotalSF'] = all_df['TotalBsmtSF'] + all_df['1stFlrSF'] + all_df['2ndFlrSF']
all_df = all_df.drop(['TotalBsmtSF', '1stFlrSF', '2ndFlrSF'], axis=1)
print(all_df.shape)

1
2
3
4

all_df['Total_Bathrooms'] = (all_df['FullBath'] + (0.5 * all_df['HalfBath']) + all_df['BsmtFullBath'] + (0.5 * all_df['BsmtHalfBath']))
all_df['Total_porch_sf'] = (all_df['OpenPorchSF'] + all_df['3SsnPorch'] + all_df['EnclosedPorch'] + all_df['ScreenPorch'])
all_df = all_df.drop(['FullBath', 'HalfBath', 'BsmtFullBath', 'BsmtHalfBath', 'OpenPorchSF', '3SsnPorch', 'EnclosedPorch', 'ScreenPorch'], axis=1)
print(all_df.shape)

• 따라서 data 정의서 먼저 봐야 한다. : data description.txt를 먼저 봐야 한다. !!! (실무에서는 없는 경우가 많다.)
• 시각화 : 각각의 data 무한작업, 도메인 공부

1
2
3
4
5
6
7
8
9
10
11

# 연도와 관련된.
num_all_vars = list(train.select_dtypes(include=[np.number]))
year_feature = []
for var in num_all_vars:
  if 'Yr' in var:
    year_feature.append(var)
  elif 'Year' in var:
    year_feature.append(var)
  else:  
    print(var, "is not related with Year")
print(year_feature)

1
2
3
4
5
6
7
8

fig, ax = plt.subplots(3, 1, figsize=(10, 6), sharex=True, sharey=True)
for i, var in enumerate(year_feature):
  if var != 'YrSold':
    ax[i].scatter(train[var], y, alpha=0.3)
    ax[i].set_title('{}'.format(var), size=15)
    ax[i].set_ylabel('SalePrice', size=15, labelpad=12.5)
plt.tight_layout()
plt.show()

1
2

all_df = all_df.drop(['YearBuilt', 'GarageYrBlt'], axis=1)
print(all_df.shape)

1
2
3
4
5
6

# 리모델링 시점으로 부터 얼마나 되었나 + 팔리는거
YearsSinceRemodel = train['YrSold'].astype(int) - train['YearRemodAdd'].astype(int)

fig, ax = plt.subplots(figsize=(10, 6))
ax.scatter(YearsSinceRemodel, y, alpha=0.3)
fig.show()

1
2
3

all_df['YearsSinceRemodel'] = all_df['YrSold'].astype(int) - all_df['YearRemodAdd'].astype(int)
all_df = all_df.drop(['YrSold', 'YearRemodAdd'], axis=1)
print(all_df.shape)

더미변수

String data (non-Numeric)

• 명목형 : 남학생, 여학생…
• 서열형(순서) : 1등급, 2등급, 3등급 (가중치, 등급숫자 등으로 바꿀 수 있다. )

• 세부적으로 customize 하는 것이 낫다.
• 명목형 series에 따라 17개의 model을 개별적으로 만들되, 하나의 모델처럼 보이게 시각화 대시보드로 만들어 줘야 합니다.

1
2

all_df['PoolArea'].value_counts()
#0과 다른값들.. 으로 되어있어서 

1
2
3
4
5
6

# 0과 1로 나누어 적용
def count_dummy(x):
  if x > 0:
    return 1
  else:
    return 0

1
2
3
4

all_df['PoolArea'] = all_df['PoolArea'].apply(count_dummy)
all_df['PoolArea'].value_counts()

# 전체 경향 등에 거의 영향을 주지 않음 

1
2

all_df['GarageArea'] = all_df['GarageArea'].apply(count_dummy)
all_df['GarageArea'].value_counts()

1
2

all_df['Fireplaces'] = all_df['Fireplaces'].apply(count_dummy)
all_df['Fireplaces'].value_counts()

Label Encoding, Ordinal Encoding, One-Hot Encoding

• Label Encoding : 종속변수에만
• Ordinal Encoding : 독립변수에만
• 써야 하지만, 개념은 같다.
• One-Hot Encoding :

1
2
3
4
5
6
7
8
9
10

from sklearn.preprocessing import LabelEncoder
import pandas as pd

temp = pd.DataFrame({'Food_Name': ['Apple', 'Chicken', 'Broccoli'], 
                     'Calories': [95, 231, 50]})

encoder = LabelEncoder()
encoder.fit(temp['Food_Name'])
labels = encoder.transform(temp['Food_Name'])
print(list(temp['Food_Name']), "==>", labels)

1
2
3
4
5
6
7
8
9
10
11

from sklearn.preprocessing import OrdinalEncoder
import pandas as pd

temp = pd.DataFrame({'Food_Name': ['Apple', 'Chicken', 'Broccoli'], 
                     'Calories': [95, 231, 50]})

encoder = OrdinalEncoder()
labels = encoder.fit_transform(temp[['Food_Name']])
print(list(temp['Food_Name']), "==>", labels.tolist())

1
2
3
4
5
6
7
8
9
10

# import pandas as pd
# temp = pd.DataFrame({'Food_Name': ['Apple', 'Chicken', 'Broccoli'], 
#                      'Calories': [95, 231, 50]})

# temp[['Food_No']] = temp.Food_Name.replace(['Chicken', 'Broccoli', 'Apple'],[1, 2, 3])

# print(temp[['Food_Name', 'Food_No']])

#ValueError: Columns must be same length as key

1
2
3
4
5
6
7
8

import pandas as pd

temp = pd.DataFrame({'Food_Name': ['Apple', 'Chicken', 'Broccoli'], 
                     'Calories': [95, 231, 50]})

temp = pd.get_dummies(temp)
print(temp)
print(temp.shape)

1
2

all_df = pd.get_dummies(all_df).reset_index(drop=True)
all_df.shape

머신러닝 모형 학습 및 평가

데이터셋 분리 및 교차 검증

1
2
3

X = all_df.iloc[:len(y), :]
X_test = all_df.iloc[len(y):, :]
X.shape, y.shape, X_test.shape

1
2
3

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0)
X_train.shape, X_test.shape, y_train.shape, y_test.shape

평가지표

MAE

1
2
3
4
5
6
7
8
9
10

import numpy as np

def mean_absolute_error(y_true, y_pred):

  error = 0
  for yt, yp in zip(y_true, y_pred):
    error = error + np.abs(yt-yp)
  
  mae = error / len(y_true)
  return mae

1
2
3
4
5
6
7
8
9
10

import numpy as np

def mean_squared_error(y_true, y_pred):

  error = 0
  for yt, yp in zip(y_true, y_pred):
    error = error + (yt - yp) ** 2
  
  mse = error / len(y_true)
  return mse

RMSE

1
2
3
4
5
6
7
8
9
10
11

import numpy as np

def root_rmse_squared_error(y_true, ypred):
  error = 0
  
  for yt, yp in zip(y_true, y_pred):
    error = error + (yt - yp) ** 2
  
  mse = error / len(y_true)
  rmse = np.round(np.sqrt(mse), 3)
  return rmse

Test1

1
2
3
4
5
6

y_true = [400, 300, 800]
y_pred = [380, 320, 777]

print("MAE:", mean_absolute_error(y_true, y_pred))
print("MSE:", mean_squared_error(y_true, y_pred))
print("RMSE:", root_rmse_squared_error(y_true, y_pred))

Test2

1
2
3
4
5
6

y_true = [400, 300, 800, 900]
y_pred = [380, 320, 777, 600]

print("MAE:", mean_absolute_error(y_true, y_pred))
print("MSE:", mean_squared_error(y_true, y_pred))
print("RMSE:", root_rmse_squared_error(y_true, y_pred))

RMSE with Sklean

1
2
3
4

from sklearn.metrics import mean_squared_error

def rmsle(y_true, y_pred):
    return np.sqrt(mean_squared_error(y_true, y_pred))

모형 정의 및 검증 평가

1
2
3
4
5
6
7
8
9
10
11
12
13
14

from sklearn.metrics import mean_squared_error
from sklearn.model_selection import KFold, cross_val_score
from sklearn.linear_model import LinearRegression

def cv_rmse(model, n_folds=5):
    cv = KFold(n_splits=n_folds, random_state=42, shuffle=True)
    rmse_list = np.sqrt(-cross_val_score(model, X, y, scoring='neg_mean_squared_error', cv=cv))
    print('CV RMSE value list:', np.round(rmse_list, 4))
    print('CV RMSE mean value:', np.round(np.mean(rmse_list), 4))
    return (rmse_list)

n_folds = 5
rmse_scores = {}
lr_model = LinearRegression()

1
2
3

score = cv_rmse(lr_model, n_folds)
print("linear regression - mean: {:.4f} (std: {:.4f})".format(score.mean(), score.std()))
rmse_scores['linear regression'] = (score.mean(), score.std())

첫번째 최종 예측 값 제출

1
2
3
4
5
6
7
8
9

from sklearn.model_selection import cross_val_predict

X = all_df.iloc[:len(y), :]
X_test = all_df.iloc[len(y):, :]
X.shape, y.shape, X_test.shape

lr_model_fit = lr_model.fit(X, y)
final_preds = np.floor(np.expm1(lr_model_fit.predict(X_test)))
print(final_preds)

1
2
3
4

submission = pd.read_csv("sample_submission.csv")
submission.iloc[:,1] = final_preds
print(submission.head())
submission.to_csv("The_first_regression.csv", index=False)

모형 알고리즘 추가

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor
from sklearn.tree import DecisionTreeRegressor
from sklearn.linear_model import LinearRegression

# LinearRegresison
lr_model = LinearRegression()

# Tree Decision 
tree_model = DecisionTreeRegressor()

# Random Forest Regressor
rf_model = RandomForestRegressor()

# Gradient Boosting Regressor
gbr_model = GradientBoostingRegressor()

1
2
3

score = cv_rmse(lr_model, n_folds)
print("linear regression - mean: {:.4f} (std: {:.4f})".format(score.mean(), score.std()))
rmse_scores['linear regression'] = (score.mean(), score.std())

1
2
3

score = cv_rmse(tree_model, n_folds)
print("Decision Tree Regressor - mean: {:.4f} (std: {:.4f})".format(score.mean(), score.std()))
rmse_scores['Decision Tree Regressor'] = (score.mean(), score.std())

1
2
3

score = cv_rmse(rf_model, n_folds)
print("RandomForest Regressor - mean: {:.4f} (std: {:.4f})".format(score.mean(), score.std()))
rmse_scores['RandomForest Regressor'] = (score.mean(), score.std())

1
2
3

score = cv_rmse(gbr_model, n_folds)
print("Gradient Boosting Regressor - mean: {:.4f} (std: {:.4f})".format(score.mean(), score.std()))
rmse_scores['Gradient Boosting Regressor'] = (score.mean(), score.std())

1
2
3
4
5
6
7
8
9
10
11
12
13
14

fig, ax = plt.subplots(figsize=(10, 6))

ax = sns.pointplot(x=list(rmse_scores.keys()), y=[score for score, _ in rmse_scores.values()], markers=['o'], linestyles=['-'], ax=ax)
for i, score in enumerate(rmse_scores.values()):
    ax.text(i, score[0] + 0.002, '{:.6f}'.format(score[0]), horizontalalignment='left', size='large', color='black', weight='semibold')

ax.set_ylabel('Score (RMSE)', size=20, labelpad=12.5)
ax.set_xlabel('Model', size=20, labelpad=12.5)
ax.tick_params(axis='x', labelsize=13.5, rotation=10)
ax.tick_params(axis='y', labelsize=12.5)
ax.set_ylim(0, 0.25)
ax.set_title('Rmse Scores of Models without Blended_Predictions', size=20)

fig.show()

• RMSE 가 적은 것이 좋다. : 예측이 잘 된 Model이라고 할 수 있다.

1
2
3
4
5
6
7
8
9
10
11

lr_model_fit = lr_model.fit(X, y)
tree_model_fit = tree_model.fit(X, y)
rf_model_fit = rf_model.fit(X, y)
gbr_model_fit = gbr_model.fit(X, y)

def blended_learning_predictions(X): 
  blended_score = (0.3 * lr_model_fit.predict(X)) + \
  (0.1 * tree_model_fit.predict(X)) + \
  (0.3 * gbr_model_fit.predict(X)) + \
  (0.3* rf_model_fit.predict(X))
  return blended_score

1
2
3
4

blended_score = rmsle(y, blended_learning_predictions(X))
rmse_scores['blended'] = (blended_score, 0)
print('RMSLE score on train data:')
print(blended_score)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

fig, ax = plt.subplots(figsize=(10, 6))

ax = sns.pointplot(x=list(rmse_scores.keys()), y=[score for score, _ in rmse_scores.values()], markers=['o'], linestyles=['-'], ax=ax)
for i, score in enumerate(rmse_scores.values()):
    ax.text(i, score[0] + 0.002, '{:.6f}'.format(score[0]), horizontalalignment='left', size='large', color='black', weight='semibold')

ax.set_ylabel('Score (RMSE)', size=20, labelpad=12.5)
ax.set_xlabel('Model', size=20, labelpad=12.5)
ax.tick_params(axis='x', labelsize=13.5, rotation=10)
ax.tick_params(axis='y', labelsize=12.5)
ax.set_ylim(0, 0.25)

ax.set_title('Rmse Scores of Models with Blended_Predictions', size=20)

fig.show()

1
2

submission.iloc[:,1] = np.floor(np.expm1(blended_learning_predictions(X_test)))
submission.to_csv("The_second_regression.csv", index=False)

Chepter 5 _파이썬 머신러닝 완벽 가이드

ref. & copyright(c) Book

회귀

• Regression: 여러개의 독립변수와 한개의 종속변수 간의 상관관계를 모델링 하는 기법

• Regression conefficients : 독립변수의 값에 영향을 미치는 회기 계수로 선형 회기 식의 기울기에 해당

• 러닝머신의 관점

  • 독립변수 : 피처
  • 종속변수 : 결정값

= > 주어진 피처와 결정값 데이터 기반에서 학습을 통해 최적의 **회귀계수** 를 찾아 내는 것이 목표

✌지도학습 2가지 유형

1. CLASSIFICATION + category, 이산 값 일때
2. REGRESSION + 숫자, 연속 값 일때

⚡ 회귀의 4가지 유형

1. 독립변수 개수 - 단일 회귀 - 다중 회귀
2. 회귀 계수의 결합 - 선형 회귀 : 실제 값과 예측 값의 차이 (오류의 제곱값)를 최소화 하는 직선형 회귀선을 최적화 하는 방식
   • Regularization(규제방법) : 일반적 선형 회귀의 과적합 문제를 해결 하기 위해 회귀 계수를 조정 하는것 (패널티 값 적용) - 비선형 회귀

Ref. scikit-learn

y = 4x + 6 + error 시뮬레이션 데이터 값 생성

1
2
3
4
5
6
7
8
9
10

import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

np.random.seed(123) # 실험 재현성

X = 2 * np.random.rand(100, 1) # 100개의 랜덤값 만들기
y = 4 * X + 6 + np.random.rand(100, 1)

plt.scatter(X, y)

1

X.shape, y.shape

((100, 1), (100, 1))

경사하강법으로 최적의 기울기 찾기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

# w1과, w0를 업데이트할 w1_update, w0_update 값 반환

def get_weight_updates(w1, w0, X, y, learning_rate=0.01):
  N = len(y)

  # w1_update, w0_update 초기화
  w1_update = np.zeros_like(w1)
  w0_update = np.zeros_like(w0)

  # 예측 배열 계산하고, 예측값과 실젯 값의 차이 계산
  y_pred = np.dot(X, w1.T) + w0
  diff = y - y_pred # 실제갓, 예측값 == 오차

  # w0_update를 dot 행렬 연산으로 구하기 위해 모두 1 값을 가진 행렬 생성
  w0_factors = np.ones((N, 1))
  
  # w1과 w0을 업데이트할 w1_update, w0_update 계산 
  w1_update = -(2/N) * learning_rate * (np.dot(X.T, diff))
  w0_update = -(2/N) * learning_rate * (np.dot(w0_factors.T, diff))

  return w1_update, w0_update

1
2
3
4
5
6
7
8
9
10
11
12
13

w0 = np.zeros((1, 1))
w1 = np.zeros((1, 1))

y_pred = np.dot(X, w1.T) + w0
diff = y-y_pred
print(diff.shape)

w0_factors = np.ones((100, 1))
w1_update = -(2/100) * 0.01 * (np.dot(X.T, diff))
w0_update = -(2/100) * 0.01 * (np.dot(w0_factors.T, diff))

print(w1_update.shape, w0_update.shape)
print(w1, w0)

(100, 1)
(1, 1) (1, 1)
[[0.]] [[0.]]

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

# 입력 인자 반복문 코드 

def gradient_descent_steps(X, y, iters = 100000):

  # w0와 w1을 모두 0으로 초기화
  w0 = np.zeros((1, 1))
  w1 = np.zeros((1, 1))

  # iters 만큼 반복 수행 # get_weight_updates
  for ind in range(iters):
    w1_update, w0_update = get_weight_updates(w1, w0, X, y, learning_rate=0.01)
    w1 = w1 - w1_update
    w0 = w0 - w0_update
  
  return w1, w0

• 예측 오차 비용 계산하는 함수 생성 및 경사 하강법 수행

1
2
3
4
5
6
7
8
9
10
11

def get_cost(y, y_pred):
  N = len(y)

  cost = np.sum(np.square(y - y_pred)) / N 
  return cost 

w1, w0 = gradient_descent_steps(X, y, iters = 100000)
print("w1:{0:.4f}, w0:{1:.4f}".format(w1[0, 0], w0[0, 0]))

y_pred = w1[0,0] * X + w0
print("Total Cost:{0:.4f}".format(get_cost(y, y_pred)))

w1:3.9462, w0:6.5590
Total Cost:0.0803

1
2

plt.scatter(X, y)
plt.plot(X, y_pred, color = "r")

1
2
3
4
5

import pandas as pd


bostonDF = pd.read_csv("https://raw.githubusercontent.com/selva86/datasets/master/BostonHousing.csv")
bostonDF.head()

EDA

• 종속변수가 기준, y값, medv

1
2
3
4
5
6
7
8
9
10
11

import matplotlib.pyplot as plt
import seaborn as sns

fig, ax = plt.subplots(figsize = (16, 8), ncols = 4, nrows = 2)
lm_features = ["rm", "zn", "indus", "nox", "age", "ptratio", "lstat", "rad"]

for i, feature in enumerate(lm_features):
  row = int(i/4)
  col = i%4
  print("row is {}, col is {}".format(row, col))
  sns.regplot(x = feature, y = "medv", data = bostonDF, ax = ax[row][col])

• 두 연속형 변수를 활용한 산점도나 회귀식 가능.
• 박스플롯 (x: 명목형, y: medv)

1
2
3
4
5
6
7

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression # model

y_target = bostonDF["medv"] # 종속변수, Y
X_data = bostonDF.drop(['medv', 'rad', 'zn', 'b', 'tax', 'age', 'indus', 'crim', 'lstat'], axis = 1, inplace = False) # 독립변수

y_target.shape, X_data.shape

((506,), (506, 5))

데이터셋 분리

• 예측, 시뮬레이션, 가상의 데이터를 가지고 예측 & 시뮬레이션
• 예측한 결괏값 vs 실젯값 비교

1
2
3
4

# 임의 샘플링

X_train, X_test, y_train, y_test = train_test_split(X_data, y_target, test_size = 0.3, random_state=156)
X_train.shape, X_test.shape, y_train.shape, y_test.shape

((354, 5), (152, 5), (354,), (152,))

ML 모형 만들기

1
2
3
4
5
6

lr = LinearRegression()
lr.fit(X_train, y_train)


y_preds = lr.predict(X_test)
y_preds

array([26.78074859, 16.40377991, 34.38443472, 19.13328473, 32.89690238,
       19.25298249, 28.32071818, 22.76654888,  9.87108567, 14.66339227,
       21.55844556, 17.27788854, 28.55574467, 38.50512646, 23.60848806,
       24.03347202, 23.82317119, 15.9119451 , 28.65132167, 20.98388455,
       20.29188703, 18.37003455, 18.58675839, 14.89143225, 35.24799305,
        7.70600921, 19.39133905, 15.97963635, 16.90296718, 15.484303  ,
       29.67753869, 17.58268684, 16.91992352, 22.47407959, 16.57706526,
       18.5381101 , 13.34337954, 24.11893098, 15.48185399, 24.3234222 ,
       36.24776797, 19.60882283, 20.95016211,  6.85667164, 20.32077896,
       23.05614583, 24.65371876, 35.25609168, 22.32959594, 25.96437918,
       27.29101785, 43.32992941, 41.76994078, 19.34288261, 24.8690423 ,
       25.99270875, 20.76285715, 33.13792328, 25.00439224, 16.82906893,
       22.80895172, 23.72489982, 24.53360315, 11.82722067, 17.55728132,
       37.43371362, 33.37256916, 25.65966256, 20.90725715, 21.09529467,
       15.22097444, 30.6234335 , 37.42143489, 26.22092177, 16.71532104,
       32.62735407, 23.41004013, 23.86575538, 18.75430877, 15.9914079 ,
       30.87778491, 16.04423898, 19.01496945, 20.04269634, 28.30832805,
       15.1948795 , 30.47430322, 33.93480059, 23.87721263, 29.7167635 ,
       29.85142798, 19.10737457, 28.49523963, 27.69846662, 25.49534489,
       24.59255802, 12.34870184, 26.65951587, 31.26197918, 17.86101862,
       27.3059424 , 18.18058484, 15.67184217, 13.17304165, 17.91281425,
       23.48894551, 24.53921273, 28.14530028, 16.05340908, 24.22120622,
       21.94517346, 26.62930956, 11.39298015, 18.53099857, 22.75407122,
       33.6679728 , 23.35342973, 20.85267956, 19.69347759, 28.12264641,
       28.56541499, 17.91759633, 27.83520695, 33.8011824 , 21.75436813,
       26.6360736 , 14.70682076, 19.99114889, 21.81029849, 31.72247354,
       21.33041025, 23.52438417, 35.55842163, 20.54294729, 38.34696416,
       19.25750865, 17.07595035, 18.31764392, 17.66658651, 23.12171447,
       19.58446231, 19.90774119, 14.84809066, 19.50652744, 38.83812958,
       15.26095952, 28.56874885, 17.62298514, 22.46794555, 23.28435884,
       18.8439135 , 31.16286012])

모형 평가

1
2
3
4

from sklearn.metrics import mean_squared_error, r2_score
mse = mean_squared_error(y_test, y_preds)

print("MSE: {0:.3f}".format(mse))

MSE: 21.369

y = 상수값 + rm 기울기 x rm의 값 +

1
2
3
4
5
6
7

import numpy as np 

print("절편 값:", lr.intercept_) # 절편 값
print("회귀 계수값", np.round(lr.coef_, 1))

coeff_df = pd.Series(data=np.round(lr.coef_, 1), index = X_data.columns)
coeff_df.sort_values(ascending=False)

절편 값: 26.830373506191982
회귀 계수값 [  4.3 -33.1   6.5  -1.1  -1.2]





rm          6.5
chas        4.3
dis        -1.1
ptratio    -1.2
nox       -33.1
dtype: float64

아직 배우지 않았지만, 유용한 기능

1

!pip install pycaret

Collecting pycaret
  Downloading pycaret-2.3.5-py3-none-any.whl (288 kB)
     |████████████████████████████████| 288 kB 32.5 MB/s 
[?25hRequirement already satisfied: joblib in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.1.0)
Requirement already satisfied: textblob in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.15.3)
Collecting pandas-profiling>=2.8.0
  Downloading pandas_profiling-3.1.0-py2.py3-none-any.whl (261 kB)
     |████████████████████████████████| 261 kB 53.7 MB/s 
[?25hCollecting pyLDAvis
  Downloading pyLDAvis-3.3.1.tar.gz (1.7 MB)
     |████████████████████████████████| 1.7 MB 42.3 MB/s 
[?25h  Installing build dependencies ... [?25l[?25hdone
  Getting requirements to build wheel ... [?25l[?25hdone
  Installing backend dependencies ... [?25l[?25hdone
    Preparing wheel metadata ... [?25l[?25hdone
Requirement already satisfied: spacy<2.4.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (2.2.4)
Collecting scikit-learn==0.23.2
  Downloading scikit_learn-0.23.2-cp37-cp37m-manylinux1_x86_64.whl (6.8 MB)
     |████████████████████████████████| 6.8 MB 58.0 MB/s 
[?25hRequirement already satisfied: ipywidgets in /usr/local/lib/python3.7/dist-packages (from pycaret) (7.6.5)
Requirement already satisfied: cufflinks>=0.17.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.17.3)
Collecting scikit-plot
  Downloading scikit_plot-0.3.7-py3-none-any.whl (33 kB)
Requirement already satisfied: yellowbrick>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.3.post1)
Requirement already satisfied: nltk in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.2.5)
Collecting umap-learn
  Downloading umap-learn-0.5.2.tar.gz (86 kB)
     |████████████████████████████████| 86 kB 6.0 MB/s 
[?25hCollecting Boruta
  Downloading Boruta-0.3-py3-none-any.whl (56 kB)
     |████████████████████████████████| 56 kB 4.7 MB/s 
[?25hRequirement already satisfied: wordcloud in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.5.0)
Requirement already satisfied: seaborn in /usr/local/lib/python3.7/dist-packages (from pycaret) (0.11.2)
Requirement already satisfied: IPython in /usr/local/lib/python3.7/dist-packages (from pycaret) (5.5.0)
Requirement already satisfied: gensim<4.0.0 in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.6.0)
Collecting lightgbm>=2.3.1
  Downloading lightgbm-3.3.1-py3-none-manylinux1_x86_64.whl (2.0 MB)
     |████████████████████████████████| 2.0 MB 47.4 MB/s 
[?25hRequirement already satisfied: pandas in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.1.5)
Collecting mlxtend>=0.17.0
  Downloading mlxtend-0.19.0-py2.py3-none-any.whl (1.3 MB)
     |████████████████████████████████| 1.3 MB 60.5 MB/s 
[?25hRequirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from pycaret) (3.2.2)
Requirement already satisfied: numpy==1.19.5 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.19.5)
Collecting pyod
  Downloading pyod-0.9.5.tar.gz (113 kB)
     |████████████████████████████████| 113 kB 58.7 MB/s 
[?25hRequirement already satisfied: plotly>=4.4.1 in /usr/local/lib/python3.7/dist-packages (from pycaret) (4.4.1)
Collecting mlflow
  Downloading mlflow-1.22.0-py3-none-any.whl (15.5 MB)
     |████████████████████████████████| 15.5 MB 50.3 MB/s 
[?25hRequirement already satisfied: scipy<=1.5.4 in /usr/local/lib/python3.7/dist-packages (from pycaret) (1.4.1)
Collecting imbalanced-learn==0.7.0
  Downloading imbalanced_learn-0.7.0-py3-none-any.whl (167 kB)
     |████████████████████████████████| 167 kB 62.1 MB/s 
[?25hCollecting kmodes>=0.10.1
  Downloading kmodes-0.11.1-py2.py3-none-any.whl (19 kB)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn==0.23.2->pycaret) (3.0.0)
Requirement already satisfied: colorlover>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (0.3.0)
Requirement already satisfied: setuptools>=34.4.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (57.4.0)
Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from cufflinks>=0.17.0->pycaret) (1.15.0)
Requirement already satisfied: smart-open>=1.2.1 in /usr/local/lib/python3.7/dist-packages (from gensim<4.0.0->pycaret) (5.2.1)
Requirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (5.1.1)
Requirement already satisfied: simplegeneric>0.8 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (0.8.1)
Requirement already satisfied: pygments in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (2.6.1)
Requirement already satisfied: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (1.0.18)
Requirement already satisfied: decorator in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (4.4.2)
Requirement already satisfied: pexpect in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (4.8.0)
Requirement already satisfied: pickleshare in /usr/local/lib/python3.7/dist-packages (from IPython->pycaret) (0.7.5)
Requirement already satisfied: jupyterlab-widgets>=1.0.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (1.0.2)
Requirement already satisfied: widgetsnbextension~=3.5.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (3.5.2)
Requirement already satisfied: ipykernel>=4.5.1 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (4.10.1)
Requirement already satisfied: nbformat>=4.2.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (5.1.3)
Requirement already satisfied: ipython-genutils~=0.2.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets->pycaret) (0.2.0)
Requirement already satisfied: tornado>=4.0 in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets->pycaret) (5.1.1)
Requirement already satisfied: jupyter-client in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets->pycaret) (5.3.5)
Requirement already satisfied: wheel in /usr/local/lib/python3.7/dist-packages (from lightgbm>=2.3.1->pycaret) (0.37.0)
Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (2.8.2)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (0.11.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (1.3.2)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->pycaret) (3.0.6)
Requirement already satisfied: jsonschema!=2.5.0,>=2.4 in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets->pycaret) (2.6.0)
Requirement already satisfied: jupyter-core in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets->pycaret) (4.9.1)
Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas->pycaret) (2018.9)
Requirement already satisfied: tqdm>=4.48.2 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (4.62.3)
Collecting tangled-up-in-unicode==0.1.0
  Downloading tangled_up_in_unicode-0.1.0-py3-none-any.whl (3.1 MB)
     |████████████████████████████████| 3.1 MB 47.3 MB/s 
[?25hRequirement already satisfied: markupsafe~=2.0.1 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (2.0.1)
Collecting pydantic>=1.8.1
  Downloading pydantic-1.8.2-cp37-cp37m-manylinux2014_x86_64.whl (10.1 MB)
     |████████████████████████████████| 10.1 MB 37.6 MB/s 
[?25hCollecting htmlmin>=0.1.12
  Downloading htmlmin-0.1.12.tar.gz (19 kB)
Collecting multimethod>=1.4
  Downloading multimethod-1.6-py3-none-any.whl (9.4 kB)
Collecting PyYAML>=5.0.0
  Downloading PyYAML-6.0-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_12_x86_64.manylinux2010_x86_64.whl (596 kB)
     |████████████████████████████████| 596 kB 40.0 MB/s 
[?25hCollecting phik>=0.11.1
  Downloading phik-0.12.0-cp37-cp37m-manylinux2010_x86_64.whl (675 kB)
     |████████████████████████████████| 675 kB 62.8 MB/s 
[?25hRequirement already satisfied: jinja2>=2.11.1 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (2.11.3)
Collecting visions[type_image_path]==0.7.4
  Downloading visions-0.7.4-py3-none-any.whl (102 kB)
     |████████████████████████████████| 102 kB 12.8 MB/s 
[?25hCollecting joblib
  Downloading joblib-1.0.1-py3-none-any.whl (303 kB)
     |████████████████████████████████| 303 kB 71.5 MB/s 
[?25hCollecting requests>=2.24.0
  Downloading requests-2.26.0-py2.py3-none-any.whl (62 kB)
     |████████████████████████████████| 62 kB 995 kB/s 
[?25hRequirement already satisfied: missingno>=0.4.2 in /usr/local/lib/python3.7/dist-packages (from pandas-profiling>=2.8.0->pycaret) (0.5.0)
Requirement already satisfied: attrs>=19.3.0 in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (21.2.0)
Requirement already satisfied: networkx>=2.4 in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (2.6.3)
Collecting imagehash
  Downloading ImageHash-4.2.1.tar.gz (812 kB)
     |████████████████████████████████| 812 kB 49.7 MB/s 
[?25hRequirement already satisfied: Pillow in /usr/local/lib/python3.7/dist-packages (from visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (7.1.2)
Collecting scipy<=1.5.4
  Downloading scipy-1.5.4-cp37-cp37m-manylinux1_x86_64.whl (25.9 MB)
     |████████████████████████████████| 25.9 MB 1.6 MB/s 
[?25hRequirement already satisfied: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly>=4.4.1->pycaret) (1.3.3)
Requirement already satisfied: wcwidth in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->IPython->pycaret) (0.2.5)
Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.7/dist-packages (from pydantic>=1.8.1->pandas-profiling>=2.8.0->pycaret) (3.10.0.2)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2.10)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2021.10.8)
Requirement already satisfied: charset-normalizer~=2.0.0 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (2.0.8)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests>=2.24.0->pandas-profiling>=2.8.0->pycaret) (1.24.3)
Requirement already satisfied: blis<0.5.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (0.4.1)
Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.6)
Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (3.0.6)
Requirement already satisfied: plac<1.2.0,>=0.9.6 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.1.3)
Requirement already satisfied: thinc==7.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (7.4.0)
Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (2.0.6)
Requirement already satisfied: wasabi<1.1.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (0.8.2)
Requirement already satisfied: catalogue<1.1.0,>=0.0.7 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.0)
Requirement already satisfied: srsly<1.1.0,>=1.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<2.4.0->pycaret) (1.0.5)
Requirement already satisfied: importlib-metadata>=0.20 in /usr/local/lib/python3.7/dist-packages (from catalogue<1.1.0,>=0.0.7->spacy<2.4.0->pycaret) (4.8.2)
Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=0.20->catalogue<1.1.0,>=0.0.7->spacy<2.4.0->pycaret) (3.6.0)
Requirement already satisfied: notebook>=4.4.1 in /usr/local/lib/python3.7/dist-packages (from widgetsnbextension~=3.5.0->ipywidgets->pycaret) (5.3.1)
Requirement already satisfied: nbconvert in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (5.6.1)
Requirement already satisfied: terminado>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.12.1)
Requirement already satisfied: Send2Trash in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (1.8.0)
Requirement already satisfied: pyzmq>=13 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel>=4.5.1->ipywidgets->pycaret) (22.3.0)
Requirement already satisfied: ptyprocess in /usr/local/lib/python3.7/dist-packages (from terminado>=0.8.1->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.7.0)
Requirement already satisfied: PyWavelets in /usr/local/lib/python3.7/dist-packages (from imagehash->visions[type_image_path]==0.7.4->pandas-profiling>=2.8.0->pycaret) (1.2.0)
Collecting docker>=4.0.0
  Downloading docker-5.0.3-py2.py3-none-any.whl (146 kB)
     |████████████████████████████████| 146 kB 70.7 MB/s 
[?25hRequirement already satisfied: entrypoints in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (0.3)
Requirement already satisfied: protobuf>=3.7.0 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (3.17.3)
Requirement already satisfied: cloudpickle in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.3.0)
Collecting databricks-cli>=0.8.7
  Downloading databricks-cli-0.16.2.tar.gz (58 kB)
     |████████████████████████████████| 58 kB 5.9 MB/s 
[?25hRequirement already satisfied: click>=7.0 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (7.1.2)
Collecting querystring-parser
  Downloading querystring_parser-1.2.4-py2.py3-none-any.whl (7.9 kB)
Collecting alembic<=1.4.1
  Downloading alembic-1.4.1.tar.gz (1.1 MB)
     |████████████████████████████████| 1.1 MB 59.1 MB/s 
[?25hCollecting prometheus-flask-exporter
  Downloading prometheus_flask_exporter-0.18.6-py3-none-any.whl (17 kB)
Requirement already satisfied: sqlalchemy in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.4.27)
Requirement already satisfied: Flask in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (1.1.4)
Collecting gitpython>=2.1.0
  Downloading GitPython-3.1.24-py3-none-any.whl (180 kB)
     |████████████████████████████████| 180 kB 58.3 MB/s 
[?25hCollecting gunicorn
  Downloading gunicorn-20.1.0-py3-none-any.whl (79 kB)
     |████████████████████████████████| 79 kB 8.7 MB/s 
[?25hRequirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (21.3)
Requirement already satisfied: sqlparse>=0.3.1 in /usr/local/lib/python3.7/dist-packages (from mlflow->pycaret) (0.4.2)
Collecting Mako
  Downloading Mako-1.1.6-py2.py3-none-any.whl (75 kB)
     |████████████████████████████████| 75 kB 4.4 MB/s 
[?25hCollecting python-editor>=0.3
  Downloading python_editor-1.0.4-py3-none-any.whl (4.9 kB)
Requirement already satisfied: tabulate>=0.7.7 in /usr/local/lib/python3.7/dist-packages (from databricks-cli>=0.8.7->mlflow->pycaret) (0.8.9)
Collecting websocket-client>=0.32.0
  Downloading websocket_client-1.2.3-py3-none-any.whl (53 kB)
     |████████████████████████████████| 53 kB 2.1 MB/s 
[?25hCollecting gitdb<5,>=4.0.1
  Downloading gitdb-4.0.9-py3-none-any.whl (63 kB)
     |████████████████████████████████| 63 kB 1.9 MB/s 
[?25hCollecting smmap<6,>=3.0.1
  Downloading smmap-5.0.0-py3-none-any.whl (24 kB)
Requirement already satisfied: greenlet!=0.4.17 in /usr/local/lib/python3.7/dist-packages (from sqlalchemy->mlflow->pycaret) (1.1.2)
Requirement already satisfied: Werkzeug<2.0,>=0.15 in /usr/local/lib/python3.7/dist-packages (from Flask->mlflow->pycaret) (1.0.1)
Requirement already satisfied: itsdangerous<2.0,>=0.24 in /usr/local/lib/python3.7/dist-packages (from Flask->mlflow->pycaret) (1.1.0)
Requirement already satisfied: mistune<2,>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.8.4)
Requirement already satisfied: bleach in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (4.1.0)
Requirement already satisfied: testpath in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.5.0)
Requirement already satisfied: defusedxml in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.7.1)
Requirement already satisfied: pandocfilters>=1.4.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (1.5.0)
Requirement already satisfied: webencodings in /usr/local/lib/python3.7/dist-packages (from bleach->nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets->pycaret) (0.5.1)
Requirement already satisfied: prometheus-client in /usr/local/lib/python3.7/dist-packages (from prometheus-flask-exporter->mlflow->pycaret) (0.12.0)
Requirement already satisfied: future in /usr/local/lib/python3.7/dist-packages (from pyLDAvis->pycaret) (0.16.0)
Collecting pyLDAvis
  Downloading pyLDAvis-3.3.0.tar.gz (1.7 MB)
     |████████████████████████████████| 1.7 MB 37.5 MB/s 
[?25h  Installing build dependencies ... [?25l[?25hdone
  Getting requirements to build wheel ... [?25l[?25hdone
  Installing backend dependencies ... [?25l[?25hdone
    Preparing wheel metadata ... [?25l[?25hdone
  Downloading pyLDAvis-3.2.2.tar.gz (1.7 MB)
     |████████████████████████████████| 1.7 MB 45.7 MB/s 
[?25hRequirement already satisfied: numexpr in /usr/local/lib/python3.7/dist-packages (from pyLDAvis->pycaret) (2.7.3)
Collecting funcy
  Downloading funcy-1.16-py2.py3-none-any.whl (32 kB)
Requirement already satisfied: numba>=0.35 in /usr/local/lib/python3.7/dist-packages (from pyod->pycaret) (0.51.2)
Requirement already satisfied: statsmodels in /usr/local/lib/python3.7/dist-packages (from pyod->pycaret) (0.10.2)
Requirement already satisfied: llvmlite<0.35,>=0.34.0.dev0 in /usr/local/lib/python3.7/dist-packages (from numba>=0.35->pyod->pycaret) (0.34.0)
Requirement already satisfied: patsy>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from statsmodels->pyod->pycaret) (0.5.2)
Collecting pynndescent>=0.5
  Downloading pynndescent-0.5.5.tar.gz (1.1 MB)
     |████████████████████████████████| 1.1 MB 49.9 MB/s 
[?25hBuilding wheels for collected packages: htmlmin, imagehash, alembic, databricks-cli, pyLDAvis, pyod, umap-learn, pynndescent
  Building wheel for htmlmin (setup.py) ... [?25l[?25hdone
  Created wheel for htmlmin: filename=htmlmin-0.1.12-py3-none-any.whl size=27098 sha256=d7dfcc5cb8473dd5eae3fcf51c538f92f876faa04e78c8b36d9c790b9fac7e10
  Stored in directory: /root/.cache/pip/wheels/70/e1/52/5b14d250ba868768823940c3229e9950d201a26d0bd3ee8655
  Building wheel for imagehash (setup.py) ... [?25l[?25hdone
  Created wheel for imagehash: filename=ImageHash-4.2.1-py2.py3-none-any.whl size=295207 sha256=8b1e1a54f9880fb8de0530e8e168811d3264000c0375d179b04677d7db738f6f
  Stored in directory: /root/.cache/pip/wheels/4c/d5/59/5e3e297533ddb09407769762985d134135064c6831e29a914e
  Building wheel for alembic (setup.py) ... [?25l[?25hdone
  Created wheel for alembic: filename=alembic-1.4.1-py2.py3-none-any.whl size=158172 sha256=3a382d7a8aa3f735be58614dc83527e0801ccb0bc893eb96cc388ee8f0a5dd91
  Stored in directory: /root/.cache/pip/wheels/be/5d/0a/9e13f53f4f5dfb67cd8d245bb7cdffe12f135846f491a283e3
  Building wheel for databricks-cli (setup.py) ... [?25l[?25hdone
  Created wheel for databricks-cli: filename=databricks_cli-0.16.2-py3-none-any.whl size=106811 sha256=ada21177391b9688188e6f778b0ec6b6001615c2b2f13bef53090805b2f183bf
  Stored in directory: /root/.cache/pip/wheels/f4/5c/ed/e1ce20a53095f63b27b4964abbad03e59cf3472822addf7d29
  Building wheel for pyLDAvis (setup.py) ... [?25l[?25hdone
  Created wheel for pyLDAvis: filename=pyLDAvis-3.2.2-py2.py3-none-any.whl size=135618 sha256=29ef50e1603fe00d18a256b833c7feddebc16ef3ac82f37f109f991b0f95b4b0
  Stored in directory: /root/.cache/pip/wheels/f8/b1/9b/560ac1931796b7303f7b517b949d2d31a4fbc512aad3b9f284
  Building wheel for pyod (setup.py) ... [?25l[?25hdone
  Created wheel for pyod: filename=pyod-0.9.5-py3-none-any.whl size=132699 sha256=851491ca675bc8eb4d9ecfb52396f362de25c1443531f442a9528c0b9b3f7b21
  Stored in directory: /root/.cache/pip/wheels/3d/bb/b7/62b60fb451b33b0df1ab8006697fba7a6a49709a629055cf77
  Building wheel for umap-learn (setup.py) ... [?25l[?25hdone
  Created wheel for umap-learn: filename=umap_learn-0.5.2-py3-none-any.whl size=82709 sha256=f4bae757148b4cf4930e495a816ecb3f6fcc3a16d1014c85ce052bb2acccb378
  Stored in directory: /root/.cache/pip/wheels/84/1b/c6/aaf68a748122632967cef4dffef68224eb16798b6793257d82
  Building wheel for pynndescent (setup.py) ... [?25l[?25hdone
  Created wheel for pynndescent: filename=pynndescent-0.5.5-py3-none-any.whl size=52603 sha256=e19d78d031a739792e30a3bf2d93865296b6eb66226835f66a8287b1330882f1
  Stored in directory: /root/.cache/pip/wheels/af/e9/33/04db1436df0757c42fda8ea6796d7a8586e23c85fac355f476
Successfully built htmlmin imagehash alembic databricks-cli pyLDAvis pyod umap-learn pynndescent
Installing collected packages: tangled-up-in-unicode, smmap, scipy, multimethod, joblib, websocket-client, visions, scikit-learn, requests, python-editor, Mako, imagehash, gitdb, querystring-parser, PyYAML, pynndescent, pydantic, prometheus-flask-exporter, phik, htmlmin, gunicorn, gitpython, funcy, docker, databricks-cli, alembic, umap-learn, scikit-plot, pyod, pyLDAvis, pandas-profiling, mlxtend, mlflow, lightgbm, kmodes, imbalanced-learn, Boruta, pycaret
  Attempting uninstall: scipy
    Found existing installation: scipy 1.4.1
    Uninstalling scipy-1.4.1:
      Successfully uninstalled scipy-1.4.1
  Attempting uninstall: joblib
    Found existing installation: joblib 1.1.0
    Uninstalling joblib-1.1.0:
      Successfully uninstalled joblib-1.1.0
  Attempting uninstall: scikit-learn
    Found existing installation: scikit-learn 1.0.1
    Uninstalling scikit-learn-1.0.1:
      Successfully uninstalled scikit-learn-1.0.1
  Attempting uninstall: requests
    Found existing installation: requests 2.23.0
    Uninstalling requests-2.23.0:
      Successfully uninstalled requests-2.23.0
  Attempting uninstall: PyYAML
    Found existing installation: PyYAML 3.13
    Uninstalling PyYAML-3.13:
      Successfully uninstalled PyYAML-3.13
  Attempting uninstall: pandas-profiling
    Found existing installation: pandas-profiling 1.4.1
    Uninstalling pandas-profiling-1.4.1:
      Successfully uninstalled pandas-profiling-1.4.1
  Attempting uninstall: mlxtend
    Found existing installation: mlxtend 0.14.0
    Uninstalling mlxtend-0.14.0:
      Successfully uninstalled mlxtend-0.14.0
  Attempting uninstall: lightgbm
    Found existing installation: lightgbm 2.2.3
    Uninstalling lightgbm-2.2.3:
      Successfully uninstalled lightgbm-2.2.3
  Attempting uninstall: imbalanced-learn
    Found existing installation: imbalanced-learn 0.8.1
    Uninstalling imbalanced-learn-0.8.1:
      Successfully uninstalled imbalanced-learn-0.8.1
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
google-colab 1.0.0 requires requests~=2.23.0, but you have requests 2.26.0 which is incompatible.
datascience 0.10.6 requires folium==0.2.1, but you have folium 0.8.3 which is incompatible.
albumentations 0.1.12 requires imgaug<0.2.7,>=0.2.5, but you have imgaug 0.2.9 which is incompatible.
Successfully installed Boruta-0.3 Mako-1.1.6 PyYAML-6.0 alembic-1.4.1 databricks-cli-0.16.2 docker-5.0.3 funcy-1.16 gitdb-4.0.9 gitpython-3.1.24 gunicorn-20.1.0 htmlmin-0.1.12 imagehash-4.2.1 imbalanced-learn-0.7.0 joblib-1.0.1 kmodes-0.11.1 lightgbm-3.3.1 mlflow-1.22.0 mlxtend-0.19.0 multimethod-1.6 pandas-profiling-3.1.0 phik-0.12.0 prometheus-flask-exporter-0.18.6 pyLDAvis-3.2.2 pycaret-2.3.5 pydantic-1.8.2 pynndescent-0.5.5 pyod-0.9.5 python-editor-1.0.4 querystring-parser-1.2.4 requests-2.26.0 scikit-learn-0.23.2 scikit-plot-0.3.7 scipy-1.5.4 smmap-5.0.0 tangled-up-in-unicode-0.1.0 umap-learn-0.5.2 visions-0.7.4 websocket-client-1.2.3

1
2

from pycaret.utils import enable_colab
enable_colab()

Colab mode enabled.

1
2

from pycaret.datasets import get_data
dataset = get_data('diamond')

1
2
3
4
5
6
7
8

data = dataset.sample(frac=0.9, random_state=786)
data_unseen = dataset.drop(data.index)

data.reset_index(drop=True, inplace=True)
data_unseen.reset_index(drop=True, inplace=True)

print('Data for Modeling: ' + str(data.shape))
print('Unseen Data For Predictions: ' + str(data_unseen.shape))

Data for Modeling: (5400, 8)
Unseen Data For Predictions: (600, 8)

1
2

from pycaret.regression import *
exp_reg101 = setup(data = data, target = 'Price', session_id=123)

시간 측정 decorator 함수 만들기

import time

1

import time

실행 시간을 확인 하는 함수 만들기

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

def timer(func):
    """ 함수 실행 시간 확인
    :param func: check할 함수 넣을거임
    :return: 걸린 시간
    """
    def wrapper(*args, **kwargs):
        #현재 시간
        time_start = time.time()

        #decorated function 불러오기
        result = func(*args, **kwargs)
        time_total = time.time() - time_start
#
        print("{}, Total time is {: .2f} sec.".format(func.__name__, time_total))

        return result
    return wrapper

decorator 함수를 이용하여 시간 확인 함수 설정, 실행

1
2
3
4
5
6
7
8

@timer
def check_time(num):
    time.sleep(num)

if __name__ == "__main__":
    check_time(1.5)

out

check_time, Total time is 1.50 sec.

관련 이론들을 아래에 적어 놓았다.

• timestamp :python에서 time은 1970년 1월 1일 0시 0분 0초를 기준으로 경과한 시간을 나타냄
• time_struct class
  • timestamp가 주어 졌을때 날짜와 시간을 알아내기 위한 API 제공

time() 함수

• 현재 timestamp 얻기

in

1
2
3

secs = time.time()
print(secs)

out

1638870356.8049076

• unix timestamp는 소수로 값을 return, 정수 부분이 초 단위.

부가적인 time 함수들

1. gmtime() : GMT 기준의 time_struct type으로 변환

   in

   1 2	tm = time.gmtime(secs) print(tm)

   out

   time.struct_time(tm_year=2021, tm_mon=12, tm_mday=7, tm_hour=9, tm_min=53, tm_sec=5, tm_wday=1, tm_yday=341, tm_isdst=0)