...

src/main/python/kr/co/swh/lecture/opensource/stock/2-analysis-krx-Linear-Regression.py

@@ -1,15 +1,30 @@
+# 오류 error: Microsoft Visual C++ 14.0 is required
+# 따라서 구글에서 검색해서 설치
+# conda create -n py36 python=3.6
+# pip3 install torch===1.3.1 torchvision===0.4.2 -f https://download.pytorch.org/whl/torch_stable.html
+
 from sqlalchemy import create_engine
 import pymysql
 import pandas as pd
 import numpy as np
 
-pymysql.install_as_MySQLdb()
-engine = create_engine("mysql+mysqldb://root:"+"tokki1127"+"@127.0.0.1/stock", encoding='utf-8')
-conn = engine.connect()
+#pymysql.install_as_MySQLdb()
+#engine = create_engine("mysql+mysqldb://root:"+"tokki1127"+"@127.0.0.1/stock", encoding='utf-8')
+#conn = engine.connect()
+
+#df = pd.read_sql_query("select * from stock_trade where 종목코드 = '000660'", conn)
+#df.name = "주가변동"
+#print(df.head())
+
+df = pd.read_csv('skhynix.csv')
+print(df.head(5))
+
+# 종목코드 left zero padding 설정
+df['종목코드'] = df['종목코드'].apply(lambda x: "{:0>6d}".format(x))
+df.head(5)
+
+
 
-df = pd.read_sql_query("select * from stock_trade where 종목코드 = '000660'", conn)
-df.name = "주가변동"
-print(df.head())
 
 
 # 인덱스 확인 (그래프의 X축 : DB로 부터 읽어온 값은 DataFrame으로 불러올 시, 0부터 숫자가 붙어있음)
@@ -37,21 +52,21 @@ for i in range(0,len(df)):
      new_data['종가'][i] = df['종가'][i]
 print(new_data.head())  # 날짜와 종가에 대한 데이터
 
-# !pip install sklearn
+# !pip install scikit-learn
 
 # fastai.structured import에서 기능 생성 add_datepart
 # !pip install fastai==0.7.0
 # 설치가 잘 되지 않음 
-#from fastai.structured import  add_datepart
-#add_datepart (new_data, 'Date')
+from fastai.structured import  add_datepart
+add_datepart (new_data, 'Date')
 # 대신 아래와 같이 코딩
 
 #new_data = new_data.copy()
-new_data['mon_fri'] = 0
-print(new_data.head())
-for i in range(0,len(new_data)):
-    if (new_data['Dayofweek'][i] == 0 or new_data['Dayofweek'][i] == 4):
-        new_data['mon_fri'][i] = 1
-    else:
-        new_data['mon_fri'][i] = 0 
-new_data.drop ( 'Elapsed', axis = 1, inplace = True) #elapsed는 타임 스탬프가됩니다
+#new_data['mon_fri'] = 0
+#print(new_data.head())
+#for i in range(0,len(new_data)):
+    #if (new_data['Dayofweek'][i] == 0 or new_data['Dayofweek'][i] == 4):
+        #new_data['mon_fri'][i] = 1
+    #else:
+        #new_data['mon_fri'][i] = 0 
+#new_data.drop ( 'Elapsed', axis = 1, inplace = True) #elapsed는 타임 스탬프가됩니다