멋쟁이 사자처럼 AI 스쿨 TIL-Mini Project 4

데이터 불러오기

import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
# tensorflow 를 tf로 불러오기
import tensorflow as tf

#pandas가 모든 열을 표시하도록 옵션을 변경합니다.
pd.set_option('display.max_columns', None)

train = pd.read_csv("train.csv",encoding="cp949")
test = pd.read_csv("test.csv",encoding="cp949")
train.shape, test.shape

((301, 58), (127, 58))

label_name = "OC"

test = test.drop(columns = "OC")

데이터확인

train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 301 entries, 0 to 300
Data columns (total 58 columns):
 #   Column              Non-Null Count  Dtype  
---  ------              --------------  -----  
 0   inst_id             301 non-null    int64  
 1   OC                  301 non-null    object 
 2   sido                301 non-null    object 
 3   sgg                 301 non-null    int64  
 4   openDate            301 non-null    int64  
 5   bedCount            296 non-null    float64
 6   instkind            300 non-null    object 
 7   revenue1            293 non-null    float64
 8   salescost1          293 non-null    float64
 9   sga1                293 non-null    float64
 10  salary1             293 non-null    float64
 11  noi1                293 non-null    float64
 12  noe1                293 non-null    float64
 13  interest1           293 non-null    float64
 14  ctax1               293 non-null    float64
 15  profit1             293 non-null    float64
 16  liquidAsset1        293 non-null    float64
 17  quickAsset1         293 non-null    float64
 18  receivableS1        293 non-null    float64
 19  inventoryAsset1     293 non-null    float64
 20  nonCAsset1          293 non-null    float64
 21  tanAsset1           293 non-null    float64
 22  OnonCAsset1         293 non-null    float64
 23  receivableL1        293 non-null    float64
 24  debt1               293 non-null    float64
 25  liquidLiabilities1  293 non-null    float64
 26  shortLoan1          293 non-null    float64
 27  NCLiabilities1      293 non-null    float64
 28  longLoan1           293 non-null    float64
 29  netAsset1           293 non-null    float64
 30  surplus1            293 non-null    float64
 31  revenue2            293 non-null    float64
 32  salescost2          293 non-null    float64
 33  sga2                293 non-null    float64
 34  salary2             293 non-null    float64
 35  noi2                293 non-null    float64
 36  noe2                293 non-null    float64
 37  interest2           293 non-null    float64
 38  ctax2               293 non-null    float64
 39  profit2             293 non-null    float64
 40  liquidAsset2        293 non-null    float64
 41  quickAsset2         293 non-null    float64
 42  receivableS2        293 non-null    float64
 43  inventoryAsset2     293 non-null    float64
 44  nonCAsset2          293 non-null    float64
 45  tanAsset2           293 non-null    float64
 46  OnonCAsset2         293 non-null    float64
 47  receivableL2        293 non-null    float64
 48  debt2               293 non-null    float64
 49  liquidLiabilities2  293 non-null    float64
 50  shortLoan2          293 non-null    float64
 51  NCLiabilities2      293 non-null    float64
 52  longLoan2           293 non-null    float64
 53  netAsset2           293 non-null    float64
 54  surplus2            293 non-null    float64
 55  employee1           291 non-null    float64
 56  employee2           288 non-null    float64
 57  ownerChange         289 non-null    object 
dtypes: float64(51), int64(3), object(4)
memory usage: 136.5+ KB

결측치 확인

# 결측치를 확인합니다.
train.isnull().sum()

inst_id                0
OC                     0
sido                   0
sgg                    0
openDate               0
bedCount               5
instkind               1
revenue1               8
salescost1             8
sga1                   8
salary1                8
noi1                   8
noe1                   8
interest1              8
ctax1                  8
profit1                8
liquidAsset1           8
quickAsset1            8
receivableS1           8
inventoryAsset1        8
nonCAsset1             8
tanAsset1              8
OnonCAsset1            8
receivableL1           8
debt1                  8
liquidLiabilities1     8
shortLoan1             8
NCLiabilities1         8
longLoan1              8
netAsset1              8
surplus1               8
revenue2               8
salescost2             8
sga2                   8
salary2                8
noi2                   8
noe2                   8
interest2              8
ctax2                  8
profit2                8
liquidAsset2           8
quickAsset2            8
receivableS2           8
inventoryAsset2        8
nonCAsset2             8
tanAsset2              8
OnonCAsset2            8
receivableL2           8
debt2                  8
liquidLiabilities2     8
shortLoan2             8
NCLiabilities2         8
longLoan2              8
netAsset2              8
surplus2               8
employee1             10
employee2             13
ownerChange           12
dtype: int64

데이터 전처리

test["openDate"] = test["openDate"].fillna(0)

train["openDate"] = pd.to_datetime(train["openDate"].astype("str"), format="%Y/%m/%d")
test["openDate"] = pd.to_datetime(test["openDate"].astype("int").astype("str"), format="%Y/%m/%d", errors="coerce")

train["open_year"] = train["openDate"].dt.year
train["open_month"] = train["openDate"].dt.month
test["open_year"] = test["openDate"].dt.year
test["open_month"] = test["openDate"].dt.month

train.drop(columns="openDate", axis=1, inplace=True)
test.drop(columns="openDate", axis=1, inplace=True)

fig, ax = plt.subplots(1, 2, figsize=(32, 8))
sns.countplot(data=train, x="open_year", ax=ax[0]).set_title("Train - Year")
sns.countplot(data=test, x="open_year", ax=ax[1]).set_title("Test - Year")
plt.show()

train["sido"].isnull().sum(), test["sido"].isnull().sum()

(0, 0)

set(train["sido"].value_counts().index) - set(test["sido"].value_counts().index)

{'gangwon', 'gwangju'}

train[train["sido"]=='jeju'].shape, test[test["sido"]=='jeju'].shape

((0, 59), (3, 58))

# ~남 ~북을 제거, ex. 충남 -> 충 / 충북 -> 충
train["sido"] = train["sido"].str.replace("nam|buk", "")
test["sido"] = test["sido"].str.replace("nam|buk", "")
# 인천과 경기를 묶어줌
train["sido"] = train["sido"].str.replace("gyeonggi|incheon", "gyeon-in")
test["sido"] = test["sido"].str.replace("gyeonggi|incheon", "gyeon-in")

<ipython-input-15-b23b3513b0b1>:2: FutureWarning: The default value of regex will change from True to False in a future version.
  train["sido"] = train["sido"].str.replace("nam|buk", "")
<ipython-input-15-b23b3513b0b1>:3: FutureWarning: The default value of regex will change from True to False in a future version.
  test["sido"] = test["sido"].str.replace("nam|buk", "")
<ipython-input-15-b23b3513b0b1>:5: FutureWarning: The default value of regex will change from True to False in a future version.
  train["sido"] = train["sido"].str.replace("gyeonggi|incheon", "gyeon-in")
<ipython-input-15-b23b3513b0b1>:6: FutureWarning: The default value of regex will change from True to False in a future version.
  test["sido"] = test["sido"].str.replace("gyeonggi|incheon", "gyeon-in")

sido_list = set(train["sido"].value_counts().index) | set(test["sido"].value_counts().index)
len(sido_list)

13

sido_list

{'busan',
 'choong',
 'daegu',
 'daejeon',
 'gangwon',
 'gwangju',
 'gyeon-in',
 'gyeong',
 'jeju',
 'jeon',
 'sejong',
 'seoul',
 'ulsan'}

fig, ax = plt.subplots(1, 2, figsize=(28, 8))
sns.countplot(data=train, x="sido", ax=ax[0]).set_title("Train - sido")
sns.countplot(data=test, x="sido", ax=ax[1]).set_title("Test - sido")
plt.show()

train.hist(bins=50,figsize=(20,20))

array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb70799a0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb7020dc0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6fd9220>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6f85610>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6f34a30>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6f60d60>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6f60e50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6f1d310>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6e78a60>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6ea4e50>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6e6b280>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6e17670>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6dc6a90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6d73e80>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6daf2b0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6d5c6d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6d09ac0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6cb6eb0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6cf12b0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6c9b9d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6c52130>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6bfc850>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6c29fa0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6bde700>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6b8ae20>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6b3f580>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6b6aca0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6b22400>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6accb20>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6a842b0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6aaf9d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6a67130>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6a10850>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb69b9f70>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb69ef6d0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb699edf0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6950550>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb68fbc70>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb68b33d0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb68ddaf0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6896250>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb683e9a0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6868160>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb681e820>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb67c9f40>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb67806a0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb67abdc0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6763520>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb670ac40>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fddb66c53a0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb66edac0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb66a5220>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6fcc250>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6e38640>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb6d17a00>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x7fddb76bd0d0>]],
      dtype=object)

sns.heatmap(train.isnull())

<matplotlib.axes._subplots.AxesSubplot at 0x7fddb4691b50>

sns.heatmap(test.isnull())

<matplotlib.axes._subplots.AxesSubplot at 0x7fddb52ba9d0>

train["OC"] = train["OC"].replace("open",1)
train["OC"] = train["OC"].replace(" close",0)

# OC (value_counts)
train["OC"].value_counts()

1    286
0     15
Name: OC, dtype: int64

train["OC"].value_counts().value_counts(normalize=True)

286    0.5
15     0.5
Name: OC, dtype: float64

# countplot
sns.countplot(data=train,x="OC")

<matplotlib.axes._subplots.AxesSubplot at 0x7fddb718fa60>

train.head()

  <script>
    const buttonEl =
      document.querySelector('#df-2ffd38eb-be1d-4e84-b043-7a9a643acafc button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-2ffd38eb-be1d-4e84-b043-7a9a643acafc');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

sns.countplot(data=train, x="open_year")
plt.show()

train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 301 entries, 0 to 300
Data columns (total 59 columns):
 #   Column              Non-Null Count  Dtype  
---  ------              --------------  -----  
 0   inst_id             301 non-null    int64  
 1   OC                  301 non-null    int64  
 2   sido                301 non-null    object 
 3   sgg                 301 non-null    int64  
 4   bedCount            296 non-null    float64
 5   instkind            300 non-null    object 
 6   revenue1            293 non-null    float64
 7   salescost1          293 non-null    float64
 8   sga1                293 non-null    float64
 9   salary1             293 non-null    float64
 10  noi1                293 non-null    float64
 11  noe1                293 non-null    float64
 12  interest1           293 non-null    float64
 13  ctax1               293 non-null    float64
 14  profit1             293 non-null    float64
 15  liquidAsset1        293 non-null    float64
 16  quickAsset1         293 non-null    float64
 17  receivableS1        293 non-null    float64
 18  inventoryAsset1     293 non-null    float64
 19  nonCAsset1          293 non-null    float64
 20  tanAsset1           293 non-null    float64
 21  OnonCAsset1         293 non-null    float64
 22  receivableL1        293 non-null    float64
 23  debt1               293 non-null    float64
 24  liquidLiabilities1  293 non-null    float64
 25  shortLoan1          293 non-null    float64
 26  NCLiabilities1      293 non-null    float64
 27  longLoan1           293 non-null    float64
 28  netAsset1           293 non-null    float64
 29  surplus1            293 non-null    float64
 30  revenue2            293 non-null    float64
 31  salescost2          293 non-null    float64
 32  sga2                293 non-null    float64
 33  salary2             293 non-null    float64
 34  noi2                293 non-null    float64
 35  noe2                293 non-null    float64
 36  interest2           293 non-null    float64
 37  ctax2               293 non-null    float64
 38  profit2             293 non-null    float64
 39  liquidAsset2        293 non-null    float64
 40  quickAsset2         293 non-null    float64
 41  receivableS2        293 non-null    float64
 42  inventoryAsset2     293 non-null    float64
 43  nonCAsset2          293 non-null    float64
 44  tanAsset2           293 non-null    float64
 45  OnonCAsset2         293 non-null    float64
 46  receivableL2        293 non-null    float64
 47  debt2               293 non-null    float64
 48  liquidLiabilities2  293 non-null    float64
 49  shortLoan2          293 non-null    float64
 50  NCLiabilities2      293 non-null    float64
 51  longLoan2           293 non-null    float64
 52  netAsset2           293 non-null    float64
 53  surplus2            293 non-null    float64
 54  employee1           291 non-null    float64
 55  employee2           288 non-null    float64
 56  ownerChange         289 non-null    object 
 57  open_year           301 non-null    int64  
 58  open_month          301 non-null    int64  
dtypes: float64(51), int64(5), object(3)
memory usage: 138.9+ KB

test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 127 entries, 0 to 126
Data columns (total 58 columns):
 #   Column              Non-Null Count  Dtype  
---  ------              --------------  -----  
 0   inst_id             127 non-null    int64  
 1   sido                127 non-null    object 
 2   sgg                 127 non-null    int64  
 3   bedCount            119 non-null    float64
 4   instkind            125 non-null    object 
 5   revenue1            125 non-null    float64
 6   salescost1          125 non-null    float64
 7   sga1                125 non-null    float64
 8   salary1             125 non-null    float64
 9   noi1                125 non-null    float64
 10  noe1                125 non-null    float64
 11  interest1           125 non-null    float64
 12  ctax1               125 non-null    float64
 13  profit1             125 non-null    float64
 14  liquidAsset1        125 non-null    float64
 15  quickAsset1         125 non-null    float64
 16  receivableS1        125 non-null    float64
 17  inventoryAsset1     125 non-null    float64
 18  nonCAsset1          125 non-null    float64
 19  tanAsset1           125 non-null    float64
 20  OnonCAsset1         125 non-null    float64
 21  receivableL1        125 non-null    float64
 22  debt1               125 non-null    float64
 23  liquidLiabilities1  125 non-null    float64
 24  shortLoan1          125 non-null    float64
 25  NCLiabilities1      125 non-null    float64
 26  longLoan1           125 non-null    float64
 27  netAsset1           125 non-null    float64
 28  surplus1            125 non-null    float64
 29  revenue2            125 non-null    float64
 30  salescost2          125 non-null    float64
 31  sga2                125 non-null    float64
 32  salary2             125 non-null    float64
 33  noi2                125 non-null    float64
 34  noe2                125 non-null    float64
 35  interest2           125 non-null    float64
 36  ctax2               125 non-null    float64
 37  profit2             125 non-null    float64
 38  liquidAsset2        125 non-null    float64
 39  quickAsset2         125 non-null    float64
 40  receivableS2        125 non-null    float64
 41  inventoryAsset2     125 non-null    float64
 42  nonCAsset2          125 non-null    float64
 43  tanAsset2           125 non-null    float64
 44  OnonCAsset2         125 non-null    float64
 45  receivableL2        125 non-null    float64
 46  debt2               125 non-null    float64
 47  liquidLiabilities2  125 non-null    float64
 48  shortLoan2          125 non-null    float64
 49  NCLiabilities2      125 non-null    float64
 50  longLoan2           125 non-null    float64
 51  netAsset2           125 non-null    float64
 52  surplus2            125 non-null    float64
 53  employee1           119 non-null    object 
 54  employee2           112 non-null    object 
 55  ownerChange         112 non-null    object 
 56  open_year           126 non-null    float64
 57  open_month          126 non-null    float64
dtypes: float64(51), int64(2), object(5)
memory usage: 57.7+ KB

train[['employee1', 'employee2']] = train[['employee1', 'employee2']].astype(float)
test['employee1'] = pd.to_numeric(test['employee1'],errors = 'coerce')
test['employee2'] = pd.to_numeric(test['employee2'],errors = 'coerce')

test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 127 entries, 0 to 126
Data columns (total 58 columns):
 #   Column              Non-Null Count  Dtype  
---  ------              --------------  -----  
 0   inst_id             127 non-null    int64  
 1   sido                127 non-null    object 
 2   sgg                 127 non-null    int64  
 3   bedCount            119 non-null    float64
 4   instkind            125 non-null    object 
 5   revenue1            125 non-null    float64
 6   salescost1          125 non-null    float64
 7   sga1                125 non-null    float64
 8   salary1             125 non-null    float64
 9   noi1                125 non-null    float64
 10  noe1                125 non-null    float64
 11  interest1           125 non-null    float64
 12  ctax1               125 non-null    float64
 13  profit1             125 non-null    float64
 14  liquidAsset1        125 non-null    float64
 15  quickAsset1         125 non-null    float64
 16  receivableS1        125 non-null    float64
 17  inventoryAsset1     125 non-null    float64
 18  nonCAsset1          125 non-null    float64
 19  tanAsset1           125 non-null    float64
 20  OnonCAsset1         125 non-null    float64
 21  receivableL1        125 non-null    float64
 22  debt1               125 non-null    float64
 23  liquidLiabilities1  125 non-null    float64
 24  shortLoan1          125 non-null    float64
 25  NCLiabilities1      125 non-null    float64
 26  longLoan1           125 non-null    float64
 27  netAsset1           125 non-null    float64
 28  surplus1            125 non-null    float64
 29  revenue2            125 non-null    float64
 30  salescost2          125 non-null    float64
 31  sga2                125 non-null    float64
 32  salary2             125 non-null    float64
 33  noi2                125 non-null    float64
 34  noe2                125 non-null    float64
 35  interest2           125 non-null    float64
 36  ctax2               125 non-null    float64
 37  profit2             125 non-null    float64
 38  liquidAsset2        125 non-null    float64
 39  quickAsset2         125 non-null    float64
 40  receivableS2        125 non-null    float64
 41  inventoryAsset2     125 non-null    float64
 42  nonCAsset2          125 non-null    float64
 43  tanAsset2           125 non-null    float64
 44  OnonCAsset2         125 non-null    float64
 45  receivableL2        125 non-null    float64
 46  debt2               125 non-null    float64
 47  liquidLiabilities2  125 non-null    float64
 48  shortLoan2          125 non-null    float64
 49  NCLiabilities2      125 non-null    float64
 50  longLoan2           125 non-null    float64
 51  netAsset2           125 non-null    float64
 52  surplus2            125 non-null    float64
 53  employee1           116 non-null    float64
 54  employee2           109 non-null    float64
 55  ownerChange         112 non-null    object 
 56  open_year           126 non-null    float64
 57  open_month          126 non-null    float64
dtypes: float64(53), int64(2), object(3)
memory usage: 57.7+ KB

train[train.columns.tolist()].head(2)

  <script>
    const buttonEl =
      document.querySelector('#df-073dd921-04d4-4d86-bae0-1b6fa8fa31eb button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-073dd921-04d4-4d86-bae0-1b6fa8fa31eb');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

train = pd.get_dummies(train[train.columns.tolist()])
test =  pd.get_dummies(test[test.columns.tolist()])

set(train.columns) - set(test.columns)

{'OC', 'instkind_dental_clinic', 'sido_gangwon', 'sido_gwangju'}

train.shape, test.shape

((301, 77), (127, 74))

feature_name = train.columns.tolist()
feature_name.remove(label_name)
feature_name.remove('sido_gangwon')
feature_name.remove('instkind_dental_clinic')
feature_name.remove('sido_gwangju')
feature_name

['inst_id',
 'sgg',
 'bedCount',
 'revenue1',
 'salescost1',
 'sga1',
 'salary1',
 'noi1',
 'noe1',
 'interest1',
 'ctax1',
 'profit1',
 'liquidAsset1',
 'quickAsset1',
 'receivableS1',
 'inventoryAsset1',
 'nonCAsset1',
 'tanAsset1',
 'OnonCAsset1',
 'receivableL1',
 'debt1',
 'liquidLiabilities1',
 'shortLoan1',
 'NCLiabilities1',
 'longLoan1',
 'netAsset1',
 'surplus1',
 'revenue2',
 'salescost2',
 'sga2',
 'salary2',
 'noi2',
 'noe2',
 'interest2',
 'ctax2',
 'profit2',
 'liquidAsset2',
 'quickAsset2',
 'receivableS2',
 'inventoryAsset2',
 'nonCAsset2',
 'tanAsset2',
 'OnonCAsset2',
 'receivableL2',
 'debt2',
 'liquidLiabilities2',
 'shortLoan2',
 'NCLiabilities2',
 'longLoan2',
 'netAsset2',
 'surplus2',
 'employee1',
 'employee2',
 'open_year',
 'open_month',
 'sido_busan',
 'sido_choong',
 'sido_daegu',
 'sido_daejeon',
 'sido_gyeon-in',
 'sido_gyeong',
 'sido_jeon',
 'sido_sejong',
 'sido_seoul',
 'sido_ulsan',
 'instkind_clinic',
 'instkind_general_hospital',
 'instkind_hospital',
 'instkind_nursing_hospital',
 'instkind_traditional_clinic',
 'instkind_traditional_hospital',
 'ownerChange_change',
 'ownerChange_same']

train = train.fillna(0)
test = test.fillna(0)

# 데이터 분할
X_train = train[feature_name]
print(X_train.shape)
X_train.head(2)

(301, 73)

  <script>
    const buttonEl =
      document.querySelector('#df-47aaf846-3fed-4922-9b36-0308f3ae6393 button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-47aaf846-3fed-4922-9b36-0308f3ae6393');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

X_test = test[feature_name]
print(X_test.shape)
X_test.head(2)

(127, 73)

<svg xmlns="http://www.w3.org/2000/svg" height="24px"viewBox="0 0 24 24"
   width="24px">
<path d="M0 0h24v24H0V0z" fill="none"/>
<path d="M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z"/><path d="M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z"/>

  <script>
    const buttonEl =
      document.querySelector('#df-0ae6d6a2-4f91-4804-84c7-d23049f0cc83 button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-0ae6d6a2-4f91-4804-84c7-d23049f0cc83');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

y_train = train[label_name]
print(y_train.shape)
y_train.head(2)

(301,)





0    1
1    1
Name: OC, dtype: int64

X_train.shape,X_test.shape,y_train.shape

((301, 73), (127, 73), (301,))

y_train.value_counts(1)

1    0.950166
0    0.049834
Name: OC, dtype: float64

# SMOTE는 Synthetic Minority Over-sampling Technique의 약자로 합성 소수자 오버샘플링 기법입니다.
# X, y를 학습하고 다시 샘플링합니다(fit_resample).
from imblearn.over_sampling import SMOTE

sm = SMOTE(random_state = 42)
X_resample, y_resample = sm.fit_resample(X_train,y_train)
X_resample.shape, y_resample.shape

((572, 73), (572,))

# X shape
X_train.shape, X_resample.shape

((301, 73), (572, 73))

# y shape
y_train.shape, y_resample.shape

((301,), (572,))

# y의 value_counts
y_train.value_counts()

1    286
0     15
Name: OC, dtype: int64

# 입력데이터 수 구하기
input_shape = X_resample.shape[1]
input_shape

73

# tf.keras.models.Sequential 로 입력-히든-출력(sigmoid) 레이어로 구성
model = tf.keras.models.Sequential([
    tf.keras.layers.Dense(units = 128, input_shape = [input_shape]),
    tf.keras.layers.Dense(128, activation='selu'),
    tf.keras.layers.Dense(128, activation='selu'),
    tf.keras.layers.Dropout(0.2),
    tf.keras.layers.Dense(128, activation='selu'),
    tf.keras.layers.Dropout(0.2),
    tf.keras.layers.Dense(128, activation='selu'),
    tf.keras.layers.Dense(1, activation='sigmoid')
])

# 모델 컴파일
model.compile(optimizer = 'adam',
              loss = 'binary_crossentropy',
              metrics=['accuracy'])

# 모델 요약
model.summary()

Model: "sequential_3"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 dense_18 (Dense)            (None, 128)               9472      
                                                                 
 dense_19 (Dense)            (None, 128)               16512     
                                                                 
 dense_20 (Dense)            (None, 128)               16512     
                                                                 
 dropout_4 (Dropout)         (None, 128)               0         
                                                                 
 dense_21 (Dense)            (None, 128)               16512     
                                                                 
 dropout_5 (Dropout)         (None, 128)               0         
                                                                 
 dense_22 (Dense)            (None, 128)               16512     
                                                                 
 dense_23 (Dense)            (None, 1)                 129       
                                                                 
=================================================================
Total params: 75,649
Trainable params: 75,649
Non-trainable params: 0
_________________________________________________________________

class PrintDot(tf.keras.callbacks.Callback):
    def on_epoch_end(self, epoch, logs):
        if epoch % 100 == 0: print('')
        print('.', end='')

early_stop = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=20)

# 학습하기
# callbacks=[early_stop, PrintDot()]
# history

history = model.fit(X_resample, y_resample, epochs=1000, validation_split=0.2,
                    callbacks=[early_stop, PrintDot()], verbose=0)

...........................................................................

<keras.callbacks.History at 0x7fddaef410d0>

# 학습결과의 history 값을 가져와서 비교하기 위해 데이터프레임으로 변환
df_hist = pd.DataFrame(history.history)
df_hist.tail()

  <script>
    const buttonEl =
      document.querySelector('#df-8b55d0f5-048b-4f9a-85c3-c5a52675e8d4 button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-8b55d0f5-048b-4f9a-85c3-c5a52675e8d4');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

# loss, accuracy 값 시각화 
df_hist[["loss","val_loss"]].plot()

<matplotlib.axes._subplots.AxesSubplot at 0x7fddaef54340>

df_hist[["accuracy","val_accuracy"]].plot()

<matplotlib.axes._subplots.AxesSubplot at 0x7fddaa9c5f70>

# 예측값을 y_pred 변수에 할당 후 재사용합니다.
y_pred = model.predict(X_test)
y_pred.shape
# 예측값을 1차원으로 변환
y_predict = y_pred.flatten()

4/4 [==============================] - 0s 4ms/step

sub = pd.read_csv("submission_sample.csv",encoding="cp949")

sub.head(2)

  <script>
    const buttonEl =
      document.querySelector('#df-6249274e-a0a9-46b2-9dd6-d25a13b1e61c button.colab-df-convert');
    buttonEl.style.display =
      google.colab.kernel.accessAllowed ? 'block' : 'none';

    async function convertToInteractive(key) {
      const element = document.querySelector('#df-6249274e-a0a9-46b2-9dd6-d25a13b1e61c');
      const dataTable =
        await google.colab.kernel.invokeFunction('convertToInteractive',
                                                 [key], {});
      if (!dataTable) return;

      const docLinkHtml = 'Like what you see? Visit the ' +
        '<a target="_blank" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'
        + ' to learn more about interactive tables.';
      element.innerHTML = '';
      dataTable['output_type'] = 'display_data';
      await google.colab.output.renderOutput(dataTable, element);
      const docLink = document.createElement('div');
      docLink.innerHTML = docLinkHtml;
      element.appendChild(docLink);
    }
  </script>
</div>

sub["OC"].shape

(127,)

sub["OC"] = y_predict

sub["OC"].value_counts()

1.0    119
0.0      8
Name: OC, dtype: int64

file_name = "hospital_4.csv"
file_name

'hospital_4.csv'

# csv 파일로 저장합니다.
sub.to_csv(file_name, index=False)