feature add : portfolio optimization

-def GMM():
-    pass
+import datetime
+import pandas as pd
+import numpy as np
+import FinanceDataReader as fdr
+from scipy.optimize import minimize
+import json
 
-def backtest():
-    pass
\ No newline at end of file
+class c_Models:
+    #Input 값으로, 자산 list, 사용자 포트폴리오 비중, 시작일, 마지막일
+    def __init__(self, assets, assets_w, start, end):
+        self.result = None
+        self.graph = None
+
+        data = pd.DataFrame()
+        # 전체 자산 data들을 가지고 온 후, 정리함
+        for asset in assets: #total_list:
+            tmp = fdr.DataReader(asset,start,end).Close
+            tmp.rename(columns={'Close': asset}, inplace=True)
+            data = pd.concat([data, tmp], axis=1)
+        
+        if data.isnull().values.any() == True: #불러온 data에 오류가 있다면
+            return "No Data",''
+
+        else:
+            data = data.resample('M').mean() #일별 데이터를 월별 데이터로 만들어줌
+            data = data.pct_change() #월별 주가 데이터를 이용해 수익률 데이터로 변환
+            data.dropna(inplace=True) #결측치 제외(첫 row)
+
+            self.data = data
+            self.assets_w = assets_w
+            self.mu = data.mean() * 12
+            self.cov = data.cov() * 12
+
+    #GMV 최적화 : 제약 조건은 비중합=1, 공매도 불가능
+    def gmv_opt(self):
+        n_assets = len(self.data.columns)
+        w0 = np.ones(n_assets) / n_assets
+        fun = lambda w: np.dot(w.T, np.dot(self.cov, w))
+        constraints = ({'type':'eq', 'fun':lambda x: np.sum(x)-1})
+        bd = ((0,1),) * n_assets
+        #cov = data.cov() * 12
+        
+        gmv = minimize(fun, w0, method = 'SLSQP', constraints=constraints, bounds=bd)
+        return gmv.x
+    
+    #Max Sharp ratio : risk free rate은 0.8%로 지정했고, 
+    def ms_opt(self):
+        n_assets = len(self.data.columns)
+        w0 = np.ones(n_assets) / n_assets
+        fun = lambda w: -(np.dot(w.T, self.mu) - 0.008) / np.sqrt(np.dot(w.T, np.dot(self.cov, w)))
+        bd = ((0,1),) * n_assets     
+        constraints = ({'type': 'eq', 'fun': lambda x:  np.sum(x) - 1})
+        maxsharp = minimize(fun, w0, method ='SLSQP', constraints=constraints, bounds=bd)
+        return maxsharp.x
+    
+    def rp_opt(self):
+        def RC(cov, w):
+            pfo_std = np.sqrt(np.dot(w.T, np.dot(self.cov, w)))
+            mrc = 1/pfo_std * (np.dot(self.cov, w))
+            rc = mrc * w
+            rc = rc / rc.sum()
+            return rc
+        
+        
+        def RP_objective(x):
+            pfo_std = np.sqrt(np.dot(x.T, np.dot(self.cov, x)))
+            mrc = 1/pfo_std * (np.dot(self.cov, x))
+            rc = mrc * x
+            rc = rc / rc.sum()
+
+            a = np.reshape(rc, (len(rc),1))
+            differs = a - a.T
+            objective = np.sum(np.square(differs))
+
+            return objective    
+        
+        n_assets = len(self.data.columns)
+        w0 = np.ones(n_assets) / n_assets
+        constraints = [{'type':'eq', 'fun': lambda x: np.sum(x) -1}]
+        bd = ((0,1),) * n_assets
+
+        rp = minimize(RP_objective, w0,  constraints=constraints, bounds = bd, method='SLSQP')
+
+        return rp.x     #, RC(self.cov, rp.x)