def __init__(self, **kwargs):
super().__init__(**kwargs)
self.cols = 2
self.add_widget(Label(text="Company Name:", font_size=30))
self.Ticker = TextInput(multiline=False, font_size=30)
self.add_widget(self.Ticker)
self.add_widget(Label(text="Start Refrence (yyyy-mm-dd):", font_size=30))
self.Start = TextInput(multiline=False, font_size=30)
self.add_widget(self.Start)
self.add_widget(Label(text="Stop Refrence(yyyy-mm-dd):", font_size=30))
self.Stop = TextInput(multiline=False, font_size=30)
self.add_widget(self.Stop)
self.Predict = Button(text="Predict", font_size=30)
self.Predict.bind(on_release=self.PredictButton)
self.add_widget(self.Predict)
self.Updates = Button(text="Version Information", font_size=30)
self.Updates.bind(on_press=self.InfoPopup)
self.add_widget(self.Updates)
self.OutputPrediction = Label(text="Decision", font_size=30)
self.add_widget(self.OutputPrediction)
self.Probability = Label(text="Probability", font_size=30)
self.add_widget(self.Probability)
try:
self.DH = DataHandler.BuySellHandler()
self.Predictor = Model.StockPredictingModel((window_size,7))
self.Predictor.Load()
except:
raise AssertionError("Load Failed")
def __init__(self, window_size):
super().__init__()
self.window_size = window_size
self.UI = tk.Tk()
self.UI.minsize(150, 200)
self.DH = DataHandler.BuySellHandler()
self.BuildUIElements()
self.Predictor = Model.StockPredictingModel((self.window_size, 7))
self.Predictor.Load()
self.UI.mainloop()
# All user editable variables will go here # Part of SMP500:['QCOM','NVDA','AMAT','PYPL','XLNX','TXN','WDC','STX','ADBE','MCHP','SWKS','FTNT','AVGO','ADSK','FISV','ADP','LRCX','CTSH','KLAC','ADI','AKAM','CDNS','PAYX','CTXS','QRVO','VRSN','FLIR','INTU','SNPS','ANSS','IPGP','V','HPE','ORCL','IBM','DXC','HPQ','MA'] tickers = ['QCOM','NVDA','TXN','ADP','IBM','INTC','AMD',"IBM","MSFT","ORCL"] start_date = '2014-01-01' end_date = '2020-4-10' window_size = 64 step_size = 16 CrossValSplits = 5 Epochs = 32 Batch_size = 32 # In[] # Prepare the Stock Data from the datahandler to be used in the NN DataProcessor = DataHandler.BuySellHandler() Unbatched_Data = DataProcessor.BuySell(tickers,start_date,end_date) # Now lets convert the data into rolling window batches BatchedX = DataProcessor.MultiMakeBatches(Unbatched_Data[0],window_size,step_size)[:-1] # Labels|'Adj Close', 'Close', 'High', 'Low', 'Open', 'Volume','Trends'| Stats = norm() BatchedY = Stats.cdf(DataProcessor.MultiMakeBatches(Unbatched_Data[1],window_size,step_size)[1:,-1,1].reshape(BatchedX.shape[0],1)) # In[] #Load the ML model Predictor = Model.StockPredictingModel(BatchedX[0].shape) Predictor.CVTrain(BatchedX,BatchedY,CrossValSplits,Epochs,Batch_size) # In[] #Test to see if it works on dataset print(Predictor.model.evaluate(BatchedX,BatchedY)) # In[] #Save The trained network
if Return_Values[0][0]>.5:
br = "Buy"
bkg = (0,1,0,1)
else:
br = "Sell"
bkg = (1,0,0,1)
print(br)
pb = Return_Values
except:
br = "Error"
bkg = (0,0,1,1)
pb = "Check Inputs"
self.OutputPrediction.text = br
self.OutputPrediction.color = bkg
self.Probability.text = str(pb[0][0])
class PredictorApp(App):
def build(self):
return PredictorPage()
cs = PredictorApp().run()
else:
DH = DataHandler.BuySellHandler()
Predictor = Model.StockPredictingModel((window_size,7))
Predictor.Load()
stocks = sys.argv[1].split(",")
print("Sotcks Entered:"+str(stocks))
Data = DH.BuySell(stocks,sys.argv[2],sys.argv[3])
for i in range(0,len(stocks)):
Model_throughput_X = Data[i][-1][-window_size:].reshape(1,window_size,7)
print(Predictor.Predict(Model_throughput_X))
