def behavior_ext(windows):
behavior_sequence = []
for window in windows:
behaviorFeature = []
records = np.array(window)
if len(records) != 0:
# print np.shape(records)
pd = pandas.DataFrame(records)
pdd = pd.describe()
# print pdd[1][0]
# for ii in range(1,4):
# for jj in range(1,8):
# behaviorFeature.append(pdd[ii][jj])
# behaviorFeature.append(pdd[0][1])
behaviorFeature.append(pdd[1][1])
behaviorFeature.append(pdd[2][1])
behaviorFeature.append(pdd[3][1])
# behaviorFeature.append(pdd[0][2])
# behaviorFeature.append(pdd[1][2])
# behaviorFeature.append(pdd[2][2])
# behaviorFeature.append(pdd[3][2])
# behaviorFeature.append(pdd[0][3])
behaviorFeature.append(pdd[1][3])
behaviorFeature.append(pdd[2][3])
behaviorFeature.append(pdd[3][3])
# behaviorFeature.append(pdd[0][4])
behaviorFeature.append(pdd[1][4])
behaviorFeature.append(pdd[2][4])
behaviorFeature.append(pdd[3][4])
# behaviorFeature.append(pdd[0][5])
behaviorFeature.append(pdd[1][5])
behaviorFeature.append(pdd[2][5])
behaviorFeature.append(pdd[3][5])
# behaviorFeature.append(pdd[0][6])
behaviorFeature.append(pdd[1][6])
behaviorFeature.append(pdd[2][6])
behaviorFeature.append(pdd[3][6])
# behaviorFeature.append(pdd[0][7])
behaviorFeature.append(pdd[1][7])
behaviorFeature.append(pdd[2][7])
behaviorFeature.append(pdd[3][7])
behavior_sequence.append(behaviorFeature)
return behavior_sequence
Spyder Editor
This is a temporary script file.
"""
#实训1
from sqlalchemy imort create_enqine
import pandas as pd
import numpy as np
engine = create_engine('data\第4章\Training_Master.csv',sep=',',encoding="gbk")
print(pd.ndim)#维度
print(pd.shape)#大小
print(pd.memory_usage())#内存信息
print(pd.describe())
def dropNullStd(data):
beforelen = data.shape[1]
colisNull = data.describe().loc['count'] == 0
for i in range(len(colisNull)):
if colisNull[i]:
data.drop(colisNull.index[i],axis=1,inplace=True)
stdisZero = data.describe().loc['std'] == 0
for i in range(len(stdisZero)):
if stdisZero[i]:
data.drop(stdisZero.index[i],axis=1,inplace=True)
afterlen = data.shape[1]
print('剔除的列的数目为：',beforelen-afterlen)
print('剔出后数据的形状为：',data.shape)
dropNullStd(pd)
# class.mro() 상속 관계를 확인할 수 있는 메서드
# 출력 [<class 'library.class>,<class 'library.class'>,<class 'object'>]
# 세줄문자: 텍스트의 한 줄이 끝남을 표시하는 문자 또는 문자열
# (개행 문자, 줄바꿈 문자, EOL과 같은 뜻)
import seaborn as sns
sns.pairplot(data,diag_kind='kde' = 커널밀도추정곡선,
diag_kind = 'hist' = 히스토그램
hue='speices',
palette = '색상')
import pandas as pd
pd.DataFrame(data) = 데이터 프레임으로 만들어줌
pd.describe() = 연산 가능한 숫자를 가진 컬럼을 추출 -> 기본통계량을 산출
what is Object Oriented Programming?(객체 지향 프로그래밍)
문제를 여러 개의 객체 단위로 나눠 작업하는 방식.
클래스를 이용해 함수(처리부분)와 변수(데이터 부분)를
하나로 묶어 객체(인스턴스)를 생성해 사용한다는 점이다.
객체: 실제 존재하는 모든 사물 또는 개념
클래스: 객체를 정의해 놓은 것
인스턴스: 객체와 비슷. 클래스로부터 객체를 만드는 과정을
'클래스의 인스턴스화'라고 부름
객체 - 핸드폰
클래스 - 핸드폰 설계도
# In[88]: pd.shape # In[89]: pd.info() # In[90]: pd.describe() # In[91]: #ADVANCED # 3 Print only the column age pd["age"] # In[93]: #ADVANCED # 4 Print only the columns age,children and charges pd[["age","children","charges"]]
df = df.append(
{
'Name': "Geetika",
'Age': 20,
'Email-id': "*****@*****.**",
'Phone-No': 8295689593
},
ignore_index=True)
print("\n", df)
#Q.2 - Download the dataset from this link ,
# Click Here
# Import the data and print the following :
# a.) First 5 rows of Dataframe
# b.) First 10 rows of the Dataframe
# c.) find basic statistics on the particular dataset.
# d.) Find the last 5 rows of the dataframe
# e.) Extract the 2nd column and find basic statistics on it.
import pandas as p
df = p.read_csv("Weather.csv")
print("First 5 rows of Dataframe:-", df.head(5))
print("First 10 rows of the Dataframe:-", df[0:10])
print("find basic statistics on the particular dataset:-", df.describe())
print("Find the last 5 rows of the dataframe:-", df.tail(5))
p = df["Location"]
print("Extract the 2nd column and find basic statistics on it:-", p.describe())
x = pca.fit_transform(x_no)
y = y_no
iteraciones = 1000
error = [None] * iteraciones
for i in range(0, iteraciones):
X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.3)
lda = LDA()
lda.fit(X_train, y_train)
error[i] = np.sum(abs(lda.predict(X_test) - y_test)) / len(y_test)
import pandas as pd
pd = pd.DataFrame(error)
print('Linear error', pd.describe(pd))
from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis as QDA
iteraciones = 1000
error = [None] * iteraciones
for i in range(0, iteraciones):
X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.3)
qda = QDA()
qda.fit(X_train, y_train)
error[i] = np.sum(abs(qda.predict(X_test) - y_test)) / len(y_test)
import pandas as pd
pd = pd.DataFrame(error)
def describe(self):
pd.describe(self.df).apply(lambda s : s.apply(lambda x : format(x, 'f')))