
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np


bike_data = pd.read_csv("bike_usage_0.csv", encoding="ansi")


bike_data.head()


labels = bike_data.Gender.dropna().unique()


sizes = bike_data.Gender.dropna().value_counts()


bike_data.Gender.value_counts()

M    11025
F     7219
m       14
f        4
Name: Gender, dtype: int64


labels

array(['M', 'F', 'm', 'f'], dtype=object)


colors = ["yellowgreen", "lightskyblue", 'lightcoral', 'blue','coral']


plt.pie(sizes, labels = labels, colors = colors, autopct = '%1.1f%%', startangle= 90)
plt.show()


plt.hist(bike_data.Distance, color='blue')
plt.show()


plt.hist(bike_data.Distance, color='blue', bins=1000)
plt.show()


plt.boxplot(bike_data.Distance)
plt.show()


under_5000 = bike_data[bike_data.Distance<5000]
plt.boxplot(under_5000.Distance)
plt.show()


under_5000 = bike_data[bike_data.Distance<5000]
plt.boxplot([under_5000.Distance[under_5000.Gender =="F"], under_5000.Distance[under_5000.Gender=="M"]])
plt.xticks([1,2],['Female','male'])
plt.show()


plt.plot(bike_data['Distance'].groupby(bike_data['Date_out']).sum())
plt.show()


plt.bar(labels, height = sizes, color ='blue')
plt.show()


bike_data['Gender'].value_counts().plot(kind='bar')
plt.show()


bike_data = pd.read_csv("bike_usage_0.csv", encoding="ansi")


bike_data.Gender.unique()

array([nan, 'M', 'F', 'm', 'f'], dtype=object)


bike_data.loc[bike_data.Gender.isnull(), 'Gender'] = "U"


bike_data.Gender.value_counts()

U    18333
M    11025
F     7219
m       14
f        4
Name: Gender, dtype: int64


bike_data[bike_data.Gender=="f"]


bike_data.loc[bike_data.Gender == 'f','Gender'] = "F"
bike_data.loc[bike_data.Gender == 'm','Gender'] = "M"


bike_data.Gender.value_counts()

U    18333
M    11039
F     7223
Name: Gender, dtype: int64


bike_data[bike_data.Distance == 0].head(2)


bike_data.loc[bike_data.Distance == 0, 'Duration'].max()

214


bike_data = bike_data[bike_data.Distance != 0]


bike_data[bike_data.Distance == 0]


len(bike_data[bike_data.Duration == 0])

0


bike_data.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 35581 entries, 0 to 36594
Data columns (total 15 columns):
 #   Column           Non-Null Count  Dtype 
---  ------           --------------  ----- 
 0   Date_out         35581 non-null  object
 1   Time_out         35581 non-null  int64 
 2   Station_no_out   35581 non-null  int64 
 3   Station_out      35581 non-null  object
 4   Membership_type  35581 non-null  object
 5   Gender           35581 non-null  object
 6   Age_Group        35581 non-null  object
 7   Momentum         35581 non-null  object
 8   Station_no_in    35581 non-null  int64 
 9   Station_in       35581 non-null  object
 10  Date_in          35581 non-null  object
 11  Bike_no          35581 non-null  object
 12  Carbon_amount    35581 non-null  object
 13  Distance         35581 non-null  int64 
 14  Duration         35581 non-null  int64 
dtypes: int64(5), object(10)
memory usage: 4.3+ MB


pd.pivot_table(bike_data, index = 'Age_Group', columns = 'Membership_type', values = "Distance", aggfunc= np.sum)


bike_pivot = pd.pivot_table(bike_data, index='Age_Group', columns='Membership_type', values='Distance', aggfunc = np.sum)


bike_pivot


bike_pivot = bike_pivot.reset_index()


pd.melt(bike_pivot, id_vars='Age_Group', value_vars=['단체권','일일권','일일권(비회원)','정기권'], var_name='Membership_type', value_name='Total_Dist')


stations = pd.read_csv("stations.csv")


stations.head()


bike_data = pd.merge(bike_data, stations, left_on='Station_no_out', right_on='ID')


bike_data.head()


bike_data.Gu.value_counts()

영등포구    10290
마포구      9345
서초구      6543
동대문구     4797
은평구      4606
Name: Gu, dtype: int64


y_gu = bike_data[bike_data.Gu=='영등포구']
m_gu = bike_data[bike_data.Gu=='마포구']


from scipy import stats
stats.levene(y_gu.Distance, m_gu.Distance)

LeveneResult(statistic=3.5647234607192013, pvalue=0.05903430224682354)


np.mean(y_gu.Distance), np.mean(m_gu.Distance)

(4190.278911564626, 4514.426966292135)


stats.ttest_ind(y_gu.Distance, m_gu.Distance, equal_var = True)

Ttest_indResult(statistic=-4.002195758414915, pvalue=6.298774059911862e-05)


s_gu = bike_data[bike_data.Gu == '서초구']
d_gu = bike_data[bike_data.Gu == '동대문구']
e_gu = bike_data[bike_data.Gu == '은평구']


from scipy import stats
stats.bartlett(y_gu.Distance, m_gu.Distance, s_gu.Distance, d_gu.Distance, e_gu.Distance)

BartlettResult(statistic=405.99591324805436, pvalue=1.4084240027307602e-86)


등분산으로 가정하고

  File "<ipython-input-48-a40b814f7648>", line 1
    등분산으로 가정하고
          ^
SyntaxError: invalid syntax


stats.f_oneway(y_gu.Distance, m_gu.Distance, s_gu.Distance, d_gu.Distance, e_gu.Distance)

F_onewayResult(statistic=37.75546101206967, pvalue=1.4366160740892166e-31)


귀무가설 기각 모든 그룹의 평균이 같지 않다.

  File "<ipython-input-50-a3c53b63c081>", line 1
    귀무가설 기각 모든 그룹의 평균이 같지 않다.
         ^
SyntaxError: invalid syntax


plot_data = [y_gu.Distance, m_gu.Distance, s_gu.Distance, d_gu.Distance, e_gu.Distance]
#plt.boxplot(plot_data)
plt.boxplot(plot_data, showfliers = False)
plt.show()


from statsmodels.stats.multicomp import pairwise_tukeyhsd

hsd = pairwise_tukeyhsd(bike_data.Distance, bike_data.Gu)
hsd.summary()


from scipy.stats import chi2_contingency

crosstab = pd.crosstab(bike_data.Age_Group, bike_data.Membership_type)
chi2_contingency(crosstab)

(1383.2239098895247,
 5.690745840063902e-283,
 18,
 array([[5.13956887e+01, 3.43679017e+03, 7.90702903e-01, 1.05780234e+04],
        [3.37559372e+01, 2.25723355e+03, 5.19322110e-01, 6.94749119e+03],
        [2.18158568e+01, 1.45880956e+03, 3.35628566e-01, 4.49003895e+03],
        [1.15089514e+01, 7.69594728e+02, 1.77060791e-01, 2.36871926e+03],
        [3.31384728e+00, 2.21594418e+02, 5.09822658e-02, 6.82040752e+02],
        [6.61308001e-01, 4.42211574e+01, 1.01739693e-02, 1.36107361e+02],
        [7.54841067e+00, 5.04756415e+02, 1.16129395e-01, 1.55357904e+03]]))


result = chi2_contingency(crosstab)
print('Chi2 Statistic :{}, p-value : {}'.format(result[0],result[1]))

Chi2 Statistic :1383.2239098895247, p-value : 5.690745840063902e-283


dist_by_gu = pd.pivot_table(bike_data, index='Gu', values ='Distance', aggfunc = len)
dist_by_gu


population = pd.read_csv('population_by_Gu.txt', sep='\t')
population


by_gu = pd.merge(dist_by_gu,population, on='Gu')
by_gu


plt.scatter(by_gu.Distance, by_gu.Population)
plt.show()


stats.pearsonr(by_gu.Distance, by_gu.Population)

(-0.3547744402265295, 0.5579504252953678)


by_gu = pd.merge(dist_by_gu, population, on ='Gu')[['Gu','Distance','Population']]
by_gu.corr()


weather = pd.read_csv("weather.csv")


weather


new_weather = pd.pivot_table(weather, index = ['date','time'], 
                             values = ['temp', 'cum_precipitation', 'humidity', 'insolation', 'sunshine', 'wind', 'wind_direction', 'sea_lvl_pressure', 'pressure'], aggfunc = np.mean)
new_weather


new_weather = new_weather.reset_index()


new_weather #분단위 데이터를 시간단위로 변경 4320 -> 72


new_bike = pd.pivot_table(bike_data, index = ['Date_out', 'Time_out'], values = ['Distance'], aggfunc = len)
new_bike = new_bike.reset_index()
new_bike


new_bike.rename(columns = {'Distance':'Count'}, inplace = True)
new_bike.columns

Index(['Date_out', 'Time_out', 'Count'], dtype='object')


bike_weather = pd.merge(new_bike, new_weather, left_on = ['Date_out', 'Time_out'], right_on = ['date','time'])
bike_weather


stats.linregress(bike_weather.temp, bike_weather.Count)

LinregressResult(slope=38.79375042397865, intercept=-320.1460800887602, rvalue=0.43840554053464026, pvalue=0.0001647148903170432, stderr=9.716288753605141)


slop, intercept, r_value, p_value, std_err = stats.linregress(bike_weather.temp, bike_weather.Count)
print("R-squared : %f" %r_value**2)

R-squared : 0.192199


import statsmodels.api as sm

x0 = bike_weather.temp
x1 = sm.add_constant(x0)
y = bike_weather.Count
model = sm.OLS(y, x1)
result = model.fit()
print(result.summary())

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Count   R-squared:                       0.192
Model:                            OLS   Adj. R-squared:                  0.180
Method:                 Least Squares   F-statistic:                     15.94
Date:                Sun, 13 Jun 2021   Prob (F-statistic):           0.000165
Time:                        09:42:49   Log-Likelihood:                -474.73
No. Observations:                  69   AIC:                             953.5
Df Residuals:                      67   BIC:                             957.9
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
==============================================================================
                 coef    std err          t      P>|t|      [0.025      0.975]
------------------------------------------------------------------------------
const       -320.1461    207.877     -1.540      0.128    -735.070      94.778
temp          38.7938      9.716      3.993      0.000      19.400      58.188
==============================================================================
Omnibus:                        9.887   Durbin-Watson:                   0.207
Prob(Omnibus):                  0.007   Jarque-Bera (JB):                3.162
Skew:                          -0.076   Prob(JB):                        0.206
Kurtosis:                       1.962   Cond. No.                         155.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.


from sklearn.model_selection import train_test_split

X = bike_weather[['cum_precipitation','humidity','temp','wind']]
y = bike_weather.Count
X_train, X_test, y_train, y_test = train_test_split(X,y, test_size =0.3, random_state =123)


import statsmodels.api as sm
X1 = sm.add_constant(X_train)
model = sm.OLS(y_train, X1)
result = model.fit()
print(result.summary())

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Count   R-squared:                       0.629
Model:                            OLS   Adj. R-squared:                  0.595
Method:                 Least Squares   F-statistic:                     18.23
Date:                Sun, 13 Jun 2021   Prob (F-statistic):           7.99e-09
Time:                        14:26:26   Log-Likelihood:                -312.08
No. Observations:                  48   AIC:                             634.2
Df Residuals:                      43   BIC:                             643.5
Df Model:                           4                                         
Covariance Type:            nonrobust                                         
=====================================================================================
                        coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------
const              1765.8805    406.043      4.349      0.000     947.017    2584.744
cum_precipitation     3.3829     12.484      0.271      0.788     -21.793      28.558
humidity            -16.5323      2.560     -6.458      0.000     -21.695     -11.370
temp                 -4.0097     11.521     -0.348      0.730     -27.244      19.224
wind                  0.9488     30.574      0.031      0.975     -60.709      62.607
==============================================================================
Omnibus:                        4.591   Durbin-Watson:                   1.995
Prob(Omnibus):                  0.101   Jarque-Bera (JB):                4.143
Skew:                           0.643   Prob(JB):                        0.126
Kurtosis:                       2.354   Cond. No.                     1.27e+03
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The condition number is large, 1.27e+03. This might indicate that there are
strong multicollinearity or other numerical problems.


X1 = sm.add_constant(X_test)
pred = result.predict(X1)
pred

8     359.689227
37    776.467633
40    596.610515
56    446.904312
23    398.884587
53    215.006054
9     468.031981
43    373.272545
68    662.458239
1     155.301651
60    600.559534
67    658.713630
42    399.725912
63    689.399962
24    351.858574
58    607.149355
59    650.659018
55    409.889723
29    371.885866
6     205.195444
61    543.214005
dtype: float64


from sklearn import metrics

print('MAE :', metrics.mean_absolute_error(y_test, pred))
print('MSE :', metrics.mean_squared_error(y_test, pred))
print('RMSE :', np.sqrt(metrics.mean_squared_error(y_test, pred)))
print('MAPE :', np.mean(np.abs((y_test -  pred) / y_test)) * 100)

MAE : 154.96781747396312
MSE : 40194.987815230634
RMSE : 200.48687691525006
MAPE : 37.74978631438906


bike_weather['Rain_YN'] ='N'
bike_weather.loc[bike_weather.cum_precipitation > 0, 'Rain_YN'] ='Y'
bike_weather


#one hot encoding


ohe  = pd.get_dummies(bike_weather['Rain_YN'])
ohe


bike_weather = pd.concat([bike_weather, ohe], axis=1, sort=False)
bike_weather


from sklearn.model_selection import train_test_split

X = bike_weather[['humidity','temp','wind','N','Y']]
y = bike_weather.Count
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size =0.3, random_state = 123)

import statsmodels.api as sm

X1 = sm.add_constant(X_train)
model = sm.OLS(y_train, X1)
result = model.fit()
print(result.summary())

X1 = sm.add_constant(X_test)
pred = result.predict(X1)

from sklearn import metrics

print('MAE :', metrics.mean_absolute_error(y_test, pred))
print('MSE :', metrics.mean_squared_error(y_test, pred))
print('RMSE :', np.sqrt(metrics.mean_squared_error(y_test, pred)))
print('MAPE :', np.mean(np.abs((y_test -  pred) / y_test)) * 100)

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Count   R-squared:                       0.639
Model:                            OLS   Adj. R-squared:                  0.606
Method:                 Least Squares   F-statistic:                     19.05
Date:                Sun, 13 Jun 2021   Prob (F-statistic):           4.44e-09
Time:                        14:42:30   Log-Likelihood:                -311.41
No. Observations:                  48   AIC:                             632.8
Df Residuals:                      43   BIC:                             642.2
Df Model:                           4                                         
Covariance Type:            nonrobust                                         
==============================================================================
                 coef    std err          t      P>|t|      [0.025      0.975]
------------------------------------------------------------------------------
const       1152.1179    259.635      4.437      0.000     628.514    1675.722
humidity     -16.2627      2.498     -6.510      0.000     -21.301     -11.225
temp          -1.6750     10.895     -0.154      0.879     -23.647      20.297
wind         -10.6456     31.830     -0.334      0.740     -74.838      53.546
N            546.2192    133.130      4.103      0.000     277.736     814.702
Y            605.8987    131.741      4.599      0.000     340.217     871.580
==============================================================================
Omnibus:                        5.222   Durbin-Watson:                   1.974
Prob(Omnibus):                  0.073   Jarque-Bera (JB):                5.093
Skew:                           0.759   Prob(JB):                       0.0784
Kurtosis:                       2.505   Cond. No.                     2.86e+17
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The smallest eigenvalue is 3.45e-30. This might indicate that there are
strong multicollinearity problems or that the design matrix is singular.
MAE : 150.83364479946903
MSE : 41933.42705995384
RMSE : 204.77652956321396
MAPE : 37.74308826445337


bike_weather['over_500'] = 1
bike_weather.loc[bike_weather.Count < 500, 'over_500'] = 0


bike_weather


from sklearn.model_selection import train_test_split

X = bike_weather[['cum_precipitation', 'humidity','temp','wind']]
y = bike_weather.over_500
X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.3, random_state=123)


import statsmodels.api as sm

X1 = sm.add_constant(X_train)
logit_model = sm.Logit(y_train, X1)
result = logit_model.fit()
print(result.summary())

Optimization terminated successfully.
         Current function value: 0.392092
         Iterations 7
                           Logit Regression Results                           
==============================================================================
Dep. Variable:               over_500   No. Observations:                   48
Model:                          Logit   Df Residuals:                       43
Method:                           MLE   Df Model:                            4
Date:                Sun, 13 Jun 2021   Pseudo R-squ.:                  0.4227
Time:                        14:48:32   Log-Likelihood:                -18.820
converged:                       True   LL-Null:                       -32.601
Covariance Type:            nonrobust   LLR p-value:                 1.530e-05
=====================================================================================
                        coef    std err          z      P>|z|      [0.025      0.975]
-------------------------------------------------------------------------------------
const                11.4773      8.488      1.352      0.176      -5.158      28.113
cum_precipitation     0.0593      0.217      0.273      0.785      -0.366       0.485
humidity             -0.1640      0.054     -3.018      0.003      -0.271      -0.057
temp                  0.0219      0.272      0.080      0.936      -0.511       0.554
wind                  0.2569      0.558      0.460      0.645      -0.837       1.350
=====================================================================================


  # Pseudo R-squ.:                  0.4227, LLR p-value:                 1.530e-05


from sklearn.linear_model import LogisticRegression

log_reg = LogisticRegression()
log_reg.fit(X_train, y_train)
print('Train set 정확도 : %.2f'%log_reg.score(X_train, y_train))
print('Test set 정확도 : %.2f'%log_reg.score(X_test,  y_test))

Train set 정확도 : 0.88
Test set 정확도 : 0.76


from sklearn.metrics import classification_report

y_pred  = log_reg.predict(X_test)
print(classification_report(y_test, y_pred))

              precision    recall  f1-score   support

           0       0.70      0.78      0.74         9
           1       0.82      0.75      0.78        12

    accuracy                           0.76        21
   macro avg       0.76      0.76      0.76        21
weighted avg       0.77      0.76      0.76        21


import os

os.environ['PATH'] += os.pathsep + 'C:\Program Files\Graphviz\bin'


#from sklearn import tree
from sklearn.tree import DecisionTreeClassifier

X = X_train
y = y_train
dTree = DecisionTreeClassifier()
dTreeModel =dTree.fit(X,y)
dTreeModel

DecisionTreeClassifier()


DecisionTreeClassifier(ccp_alpha =0.0, class_weight=None, criterion='gini',
                      max_depth =None, max_features=None, max_leaf_nodes=None,
                      min_impurity_decrease=0.0, min_impurity_split=None,
                      min_samples_leaf=1, min_samples_split=2,
                      min_weight_fraction_leaf=0.0, presort='deprecated',
                      random_state=None, splitter='best')

DecisionTreeClassifier()


from sklearn.tree import export_graphviz
import pydotplus
from IPython.display import Image

dot_data = export_graphviz(dTreeModel, out_file=None,
                          feature_names=['cum_precipitation','humidity','temp','wind'],
                          class_names =('Y','N'), filled =True, rounded=True, special_characters=True)

graph = pydotplus.graph_from_dot_data(dot_data)
Image(graph.create_png())


#지니 불순도 불순도가 낮아지도록


dTreeModel.predict(X_test)

array([1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 1],
      dtype=int64)


from sklearn.metrics import accuracy_score

y_pred = dTreeModel.predict(X_test)
print('Accuracy : %2f'%accuracy_score(y_test,y_pred))

Accuracy : 0.809524


y_pred = dTreeModel.predict(X_test)
print(classification_report(y_test,y_pred))

              precision    recall  f1-score   support

           0       0.86      0.67      0.75         9
           1       0.79      0.92      0.85        12

    accuracy                           0.81        21
   macro avg       0.82      0.79      0.80        21
weighted avg       0.82      0.81      0.80        21


# 의사결정나무 과적합이 많이 나옴
# 학습정확도 높고, 예측정확도는 낮은편
# 직관적 이해쉽고, 데이터에 대한 제약이 적다.


import pandas as pd
weather = pd.read_csv('weather.csv')
weather


from  sklearn.preprocessing import MinMaxScaler
import pandas as pd
import numpy as np

weather = pd.read_csv('weather.csv')
X = np.array(weather.humidity).reshape(-1,1)
scaler = MinMaxScaler()
X_scaled = scaler.fit_transform(X)
X_scaled

array([[0.97205589],
       [0.97205589],
       [0.97205589],
       ...,
       [0.36526946],
       [0.36926148],
       [0.37125749]])


n_bike = pd.pivot_table(bike_data, index=['Gu','Date_out','Time_out'], values = 'Distance', aggfunc = len)
n_bike = n_bike.reset_index()
n_bike.rename(columns = {'Distance':'Count'}, inplace=True)
n_bike


n_bike2 = pd.pivot_table(n_bike, index = 'Gu', columns = 'Time_out', values = 'Count', aggfunc = np.mean)
n_bike2 = n_bike2.reset_index()
n_bike2


from sklearn import cluster

X = n_bike2.iloc[0:5, 1:25]
y = n_bike2.Gu

km2 = cluster.KMeans(n_clusters=2).fit(X)
km3 = cluster.KMeans(n_clusters=3).fit(X)
km4 = cluster.KMeans(n_clusters=4).fit(X)


n_bike2['2_Cluster'] = km2.labels_
n_bike2['3_Cluster'] = km3.labels_
n_bike2['4_Cluster'] = km4.labels_
n_bike2[['Gu','2_Cluster','3_Cluster','4_Cluster']]

	Date_out	Station_no_out	Station_out	Membership_type	Gender	Age_Group	Momentum	Station_no_in	Station_in	Date_in	Bike_no	Carbon_amount	Distance	Duration
0	2019-10-03	129	129. 신촌역(2호선) 6번출구 옆	정기권	NaN	~10대	28.27	122	신성기사식당 앞	2019-10-03 오전 12:20:42	SPB-15000	0.24	1050	5
1	2019-10-03	150	150. 서강대역 2번출구 앞	정기권	NaN	20대	146.46	2065	서울시여성가족재단	2019-10-03 오전 1:16:32	SPB-13087	1.32	5690	32
2	2019-10-03	240	240. 문래역 4번출구 앞	정기권	NaN	20대	37.13	245	삼성생명 당산사옥 앞	2019-10-03 오전 12:18:21	SPB-23229	0.29	1250	10
3	2019-10-03	623	623. 서울시립대 정문 앞	정기권	NaN	20대	134.62	1346	길음8골어린이공원 옆	2019-10-03 오전 1:15:39	SPB-14181	1.21	5230	24
4	2019-10-03	633	633. 청량리 기업은행 앞	정기권	NaN	20대	85.83	568	청계8가사거리 부근	2019-10-03 오전 12:17:58	SPB-15221	0.67	2890	11

	Date_out	Time_out	Station_no_out	Station_out	Membership_type	Gender	Age_Group	Momentum	Station_no_in	Station_in	Date_in	Bike_no	Carbon_amount	Distance	Duration
16548	2019-10-04	13	205	205. 산업은행 앞	일일권	f	20대	156.29	1834	월드메르디앙 벤처센터 2차	2019-10-04 오후 2:18:10	SPB-22306	1.76	7590	59
18978	2019-10-04	17	2252	2252. 하이브랜드 앞	정기권	f	20대	63.84	2243	서울가정법원	2019-10-04 오후 5:39:57	SPB-22209	0.58	2480	13
31071	2019-10-05	15	152	152. 마포구민체육센터 앞	일일권	f	30대	132.07	106	합정역 7번출구 앞	2019-10-05 오후 5:39:38	SPB-20461	1.55	6670	113
33075	2019-10-05	18	2251	2251. 더케이호텔 입구(양재2)	정기권	f	40대	40.51	2243	서울가정법원	2019-10-05 오후 6:29:34	SPB-05988	0.45	1930	11

Membership_type	단체권	일일권	일일권(비회원)	정기권
Age_Group
20대	371870.0	24609660.0	14160.0	32250310.0
30대	92580.0	15251200.0	NaN	26300820.0
40대	177820.0	6991400.0	NaN	20321980.0
50대	39560.0	2371560.0	NaN	11612290.0
60대	NaN	388380.0	NaN	3840400.0
70대~	2450.0	76880.0	NaN	629750.0
~10대	50280.0	4623160.0	NaN	4168920.0

Membership_type	단체권	일일권	일일권(비회원)	정기권
Age_Group
20대	371870.0	24609660.0	14160.0	32250310.0
30대	92580.0	15251200.0	NaN	26300820.0
40대	177820.0	6991400.0	NaN	20321980.0
50대	39560.0	2371560.0	NaN	11612290.0
60대	NaN	388380.0	NaN	3840400.0
70대~	2450.0	76880.0	NaN	629750.0
~10대	50280.0	4623160.0	NaN	4168920.0

	Age_Group	Membership_type	Total_Dist
0	20대	단체권	371870.0
1	30대	단체권	92580.0
2	40대	단체권	177820.0
3	50대	단체권	39560.0
4	60대	단체권	NaN
5	70대~	단체권	2450.0
6	~10대	단체권	50280.0
7	20대	일일권	24609660.0
8	30대	일일권	15251200.0
9	40대	일일권	6991400.0
10	50대	일일권	2371560.0
11	60대	일일권	388380.0
12	70대~	일일권	76880.0
13	~10대	일일권	4623160.0
14	20대	일일권(비회원)	14160.0
15	30대	일일권(비회원)	NaN
16	40대	일일권(비회원)	NaN
17	50대	일일권(비회원)	NaN
18	60대	일일권(비회원)	NaN
19	70대~	일일권(비회원)	NaN
20	~10대	일일권(비회원)	NaN
21	20대	정기권	32250310.0
22	30대	정기권	26300820.0
23	40대	정기권	20321980.0
24	50대	정기권	11612290.0
25	60대	정기권	3840400.0
26	70대~	정기권	629750.0
27	~10대	정기권	4168920.0

Pivot table¶

0.05 보다 그면 등분산 작으면 등분산이 아님 0.05보다 크기때문에 등분산 가정¶

귀무가설 평균이 같다¶

0.05 보다 작기 때문에 귀무가설 기각 평균이 같지 않다¶

1. 등분산분석¶

2. 분산분석¶

3. 사후분석¶

사후분석 Tukey's HSD Test¶

영등포 은평구만 같고 나머지는 같지 않다.¶

12. 카이제곱 검정¶

13. 상관분석¶

14. 회귀분석¶

rvalue, 결정계수, 설명력¶

15. 머신러닝¶

16.로지스틱 회귀분석¶

17. 의사결정나무¶

18. 머신러닝: 군집분석¶

	Date_out	Time_out	Station_no_out	Station_out	Membership_type	Gender	Age_Group	Momentum	Station_no_in	Station_in	Date_in	Bike_no	Carbon_amount	Distance	Duration
132	2019-10-03	2	416	416. 상암월드컵파크 1단지 교차로	정기권	U	50대	0	133	해담는다리	2019-10-03 오전 3:05:25	SPB-14937	0	0	12
164	2019-10-03	2	113	113. 홍대입구역 2번출구 앞	일일권	U	20대	0	176	명지대학교 도서관	2019-10-03 오전 2:55:19	SPB-00842	0	0	27

	Gu	ID	Station	Address	Latitude	Longitude	Date	No_of_Racks
0	마포구	101	101. (구)합정동 주민센터	서울특별시 마포구 동교로8길 58	37.549561	126.905754	2015-09-06 23:40	5
1	마포구	102	102. 망원역 1번출구 앞	서울특별시 마포구 월드컵로 72	37.555649	126.910629	2015-09-06 23:42	20
2	마포구	103	103. 망원역 2번출구 앞	서울특별시 마포구 월드컵로 79	37.554951	126.910835	2015-09-06 23:43	14
3	마포구	104	104. 합정역 1번출구 앞	서울특별시 마포구 양화로 59	37.550629	126.914986	2015-09-06 23:44	13
4	마포구	105	105. 합정역 5번출구 앞	서울특별시 마포구 양화로 48	37.550007	126.914825	2015-09-06 23:45	5

group1	group2	meandiff	p-adj	lower	upper	reject
동대문구	마포구	793.049	0.001	515.079	1071.019	True
동대문구	서초구	1210.5524	0.001	913.0775	1508.0272	True
동대문구	영등포구	468.901	0.001	195.2948	742.5072	True
동대문구	은평구	351.3685	0.0249	28.5165	674.2204	True
마포구	서초구	417.5033	0.001	165.2287	669.778	True
마포구	영등포구	-324.1481	0.001	-547.7807	-100.5154	True
마포구	은평구	-441.6805	0.001	-723.4333	-159.9278	True
서초구	영등포구	-741.6514	0.001	-989.1095	-494.1933	True
서초구	은평구	-859.1839	0.001	-1160.1964	-558.1714	True
영등포구	은평구	-117.5325	0.7495	-394.9809	159.9159	False

	Distance
Gu
동대문구	4797
마포구	9345
서초구	6543
영등포구	10290
은평구	4606

	Gu	Family	Population	Male	Female	D_Total	D_Male	D_Female	F_Total	F_Male	F_Female	per_Family	over_65
0	동대문구	164191	363023	178490	184533	346194	172113	174081	16829	6377	10452	2.11	59350
1	은평구	207681	484546	233360	251186	480032	231528	248504	4514	1832	2682	2.31	80738
2	마포구	175023	385925	181303	204622	374035	176891	197144	11890	4412	7478	2.14	52429
3	영등포구	174806	400986	200986	200000	367678	182438	185240	33308	18548	14760	2.10	57872
4	서초구	173199	435107	208181	226926	430826	206039	224787	4281	2142	2139	2.49	57136

	date_old	date	time	temp	cum_precipitation	humidity	insolation	sunshine	wind	wind_direction	sea_lvl_pressure	pressure
0	2019-10-03 오전 12:00:00	2019-10-03	12	20.0	23.2	94.1	3.40	0	5.5	351.7	1004.1	994.2
1	2019-10-03 오전 12:01:00	2019-10-03	12	20.1	0.0	94.1	0.00	0	3.7	348.6	1004.1	994.2
2	2019-10-03 오전 12:02:00	2019-10-03	12	20.0	0.0	94.1	0.00	0	3.6	346.4	1004.1	994.2
3	2019-10-03 오전 12:03:00	2019-10-03	12	20.0	0.0	94.1	0.00	0	3.1	349.1	1004.1	994.2
4	2019-10-03 오전 12:04:00	2019-10-03	12	20.0	0.0	94.0	0.00	0	3.4	335.9	1004.1	994.2
...	...	...	...	...	...	...	...	...	...	...	...	...
4315	2019-10-05 오후 11:55:00	2019-10-05	23	14.8	0.6	63.6	8.24	8400	3.2	350.9	1021.3	1011.0
4316	2019-10-05 오후 11:56:00	2019-10-05	23	14.8	0.6	63.5	8.24	8400	2.9	354.3	1021.3	1011.0
4317	2019-10-05 오후 11:57:00	2019-10-05	23	14.8	0.6	63.7	8.24	8400	3.1	3.9	1021.3	1011.0
4318	2019-10-05 오후 11:58:00	2019-10-05	23	14.7	0.6	63.9	8.24	8400	2.3	10.0	1021.2	1011.0
4319	2019-10-05 오후 11:59:00	2019-10-05	23	14.8	0.6	64.0	8.24	8400	2.6	351.7	1021.3	1011.0

		cum_precipitation	humidity	insolation	pressure	sea_lvl_pressure	sunshine	temp	wind	wind_direction
date	time
2019-10-03	1	2.361667	93.846667	0.0000	993.010000	1002.910000	0	20.016667	3.290000	178.788333
	2	3.353333	93.453333	0.0000	992.668333	1002.568333	0	19.908333	3.056667	333.400000
	3	3.930000	91.686667	0.0000	992.253333	1002.153333	0	19.923333	2.125000	330.110000
	4	4.423333	93.061667	0.0000	992.316667	1002.216667	0	19.928333	1.931667	251.535000
	5	4.500000	95.028333	0.0000	992.835000	1002.735000	0	19.871667	2.886667	236.116667
...	...	...	...	...	...	...	...	...	...	...
2019-10-05	20	0.600000	61.381667	8.2400	1009.146667	1019.273333	8400	16.610000	3.631667	125.785000
	21	0.600000	62.598333	8.2400	1009.773333	1019.973333	8400	15.988333	3.686667	225.700000
	22	0.600000	63.560000	8.2400	1010.376667	1020.576667	8400	15.436667	3.680000	225.253333
	23	0.600000	63.426667	8.2400	1010.846667	1021.085000	8400	14.986667	3.401667	179.688333
	24	0.600000	69.291667	3.6575	1005.155000	1015.140000	1703	20.473333	2.841667	248.531667

Time_out	Gu	0	1	2	3	4	5	6	7	8	...	14	15	16	17	18	19	20	21	22	23
0	동대문구	47.000000	45.000000	37.333333	29.666667	15.333333	16.000000	23.666667	39.666667	61.666667	...	85.333333	84.666667	92.666667	106.666667	111.666667	99.666667	93.000000	93.333333	84.333333	83.000000
1	마포구	86.666667	70.000000	48.666667	38.666667	31.666667	27.333333	33.666667	68.666667	91.333333	...	160.333333	200.333333	195.333333	233.000000	243.666667	204.333333	199.333333	201.333333	158.333333	118.000000
2	서초구	49.000000	47.000000	33.666667	32.666667	18.666667	16.333333	26.333333	47.333333	77.333333	...	123.000000	135.000000	142.666667	175.000000	187.333333	148.000000	145.333333	131.666667	113.000000	80.333333
3	영등포구	86.333333	66.333333	49.333333	36.000000	26.333333	31.000000	53.333333	89.000000	127.333333	...	194.000000	207.333333	219.333333	249.666667	275.666667	221.333333	200.000000	216.333333	187.333333	139.333333
4	은평구	48.666667	42.000000	35.333333	23.333333	12.333333	17.666667	26.000000	45.000000	65.666667	...	85.333333	86.000000	91.000000	107.000000	97.333333	89.333333	93.666667	72.000000	79.333333	70.333333

	Distance	Population
Distance	1.000000	-0.354774
Population	-0.354774	1.000000