Spaces:
Runtime error
Runtime error
File size: 2,990 Bytes
20c3d48 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
from stocks import *
from functions import *
from datetime import datetime
import streamlit as st
st.set_page_config(layout="wide")
st.title("Tech Stocks Trading Assistant")
left_column, right_column = st.columns(2)
with left_column:
all_tickers = {
"Apple":"AAPL",
"Microsoft":"MSFT",
"Nvidia":"NVDA",
"Paypal":"PYPL",
"Amazon":"AMZN",
"Spotify":"SPOT",
"Twitter":"TWTR",
"AirBnB":"ABNB",
"Uber":"UBER",
"Google":"GOOG"
}
st.subheader("Technical Analysis Methods")
option_name = st.selectbox('Choose a stock:', all_tickers.keys())
option_ticker = all_tickers[option_name]
execution_timestamp = datetime.now()
'You selected: ', option_name, "(",option_ticker,")"
'Last execution:', execution_timestamp
s = Stock_Data()
t = s.Ticker(tick=option_ticker)
m = Models()
with st.spinner('Loading stock data...'):
technical_analysis_methods_outputs = {
'Technical Analysis Method': [
'Bollinger Bands (20 days & 2 stand. deviations)',
'Bollinger Bands (10 days & 1.5 stand. deviations)',
'Bollinger Bands (50 days & 3 stand. deviations)',
'Moving Average Convergence Divergence (MACD)'
],
'Outlook': [
m.bollinger_bands_20d_2std(t),
m.bollinger_bands_10d_1point5std(t),
m.bollinger_bands_50d_3std(t),
m.MACD(t)
],
'Timeframe of Method': [
"Medium-term",
"Short-term",
"Long-term",
"Short-term"
]
}
df = pd.DataFrame(technical_analysis_methods_outputs)
def color_survived(val):
color = ""
if (val=="Sell" or val=="Downtrend and sell signal" or val=="Downtrend and no signal"):
color="#EE3B3B"
elif (val=="Buy" or val=="Uptrend and buy signal" or val=="Uptrend and no signal"):
color="#3D9140"
else:
color="#CD950C"
return f'background-color: {color}'
st.table(df.sort_values(['Timeframe of Method'], ascending=False).
reset_index(drop=True).style.applymap(color_survived, subset=['Outlook']))
with right_column:
st.subheader("FinBERT-based Sentiment Analysis")
with st.spinner("Connecting with www.marketwatch.com..."):
st.plotly_chart(m.finbert_headlines_sentiment(t)["fig"])
"Current sentiment:", m.finbert_headlines_sentiment(t)["current_sentiment"], "%"
st.subheader("LSTM-based 7-day stock price prediction model")
with st.spinner("Compiling LSTM model.."):
st.plotly_chart(m.LSTM_7_days_price_predictor(t))
|