app.py

import requests
import json
from requests.api import delete, options
import streamlit as st
import time
import tweepy
import requests
from io import BytesIO
import base64
import matplotlib.pyplot as plt
import numpy as np
from plotly.subplots import make_subplots
from config import *
from dateutil.relativedelta import relativedelta
from patterns import patterns
# import talibsddsfs
from datetime import datetime, timedelta, tzinfo
from alpaca_trade_api.rest import REST
from streamlit_tags import st_tags_sidebar
# from streamlit_autorefresh import st_autorefresh
# import plotly.express as px
from coinbaskets import *
import plotly.graph_objects as go
from mapping import *
import pandas as pd
import threading
from bs4 import BeautifulSoup
from ETFs import *
from dateutil import tz
import os
# try:
#     from streamlit.ReportThread import add_report_ctx
# except Exception:
#     # Streamlit >= 0.65.0
#     from streamlit.report_thread import add_report_ctx
#     # from streamlit.scriptrunner import add_script_run_ctx
from streamlit.scriptrunner import add_script_run_ctx as add_report_ctx
def get_stocktwits_data(req,code,label):

            r = requests.get(req)
            trending_syms = pd.DataFrame(r.json()["stocks"]).T
            trending_syms.index.name = "stock_id"
            trending_syms.index = trending_syms.index.astype("int")
            trending_score = pd.DataFrame(r.json()["table"][code])
            trending_score.set_index("stock_id",inplace = True)
            most_trending_syms = pd.merge(trending_syms,trending_score,on= "stock_id")
            most_trending_syms.sort_values("val",ascending = False, inplace = True)
            most_trending_syms.set_index("symbol",inplace = True)
            most_trending_syms.columns = ["Name","Price","%Change",label]
            return most_trending_syms
        
def get_cnbc_data(symbol):
    ticker = symbol.replace(" ","")
    if ticker == "NASDAQ":
        ticker = "NDX"
    elif ticker == "NIFTY50":
        ticker = ".NSEI"
    # Get the symbol quote from yahoo finance, we are using Beautiful Soup for scraping
    # df = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/1Y.json?symbol={ticker}").json()["barData"]["priceBars"])
    df_1D = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/1D.json?symbol={ticker}").json()["barData"]["priceBars"])
    # df["datetime"] = pd.to_datetime(df['tradeTimeinMills'],unit='ms')
    # df["close"] = df["close"].astype(float)
    df_1D["close"] = df_1D["close"].astype(float)
    # df.set_index("datetime",inplace = True)
    # dma200 = df["close"].rolling(200).mean()
    close = (df_1D["close"].iloc[-1])
    return close
def vix_gradient(vix):
    """
    Mapping is done as follows rsi<=20 --> -100, rsi>=80, 100, and then linear variation
    """
    if vix<20:
        return 100
    elif vix<30:
        return (-20*vix+500)
    else:
        return -100
def roro_comp_get(series,i,state, inverse = False):
    
    current = series.iloc[-i]
    current_idx = series.index[-i]
        
    w_1_ago = current_idx - relativedelta(days=7)
    w_2_ago = current_idx - relativedelta(days=14)
    m_1_ago = current_idx - relativedelta(months=1)
        
    if state == 0:
        w_1_ret = (current - series.loc[w_1_ago:].iloc[0])*100/series.loc[w_1_ago:].iloc[0]
        w_2_ret = (current - series.loc[w_2_ago:].iloc[0])*100/series.loc[w_2_ago:].iloc[0]
        m_1_ret = (current - series.loc[m_1_ago:].iloc[0])*100/series.loc[m_1_ago:].iloc[0]
    else:
        w_1_ret = (current - series.iloc[-1-i])*100/series.iloc[-1-i]
        w_2_ret = (current - series.iloc[-2-i])*100/series.iloc[-2-i]
        m_1_ret = (current - series.iloc[-4-i])*100/series.iloc[-4-i]
    sign_of = 1
    if inverse  == True:
        sign_of = -1
    val = 100*(3*(2*(sign_of*w_1_ret>0) - 1) + 2*(2*(sign_of*w_2_ret>0) - 1) + 2*(sign_of*m_1_ret>0) - 1)/6
    
    return val
def get_roro(tf = "1Y"):
    
    df_spx = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=.SPX").json()["barData"]["priceBars"])
    df_spx["datetime"] = pd.to_datetime(df_spx['tradeTimeinMills'],unit='ms').dt.date
    df_spx.set_index("datetime",inplace = True)
    df_spx["close"] = df_spx["close"].astype(float)

    df_vix = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=.VIX").json()["barData"]["priceBars"])
    df_vix["close"] = df_vix["close"].astype(float)

    df_AUDJPY = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=AUDJPY=").json()["barData"]["priceBars"])
    df_AUDJPY["datetime"] = pd.to_datetime(df_AUDJPY['tradeTimeinMills'],unit='ms').dt.date
    df_AUDJPY.set_index("datetime",inplace = True)
    df_AUDJPY["close"] = df_AUDJPY["close"].astype(float)
    
   
    df_gold = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=@GC.1").json()["barData"]["priceBars"])
    df_gold["datetime"] = pd.to_datetime(df_gold['tradeTimeinMills'],unit='ms').dt.date
    df_gold.set_index("datetime",inplace = True)
    df_gold["close"] = df_gold["close"].astype(float)
    
    df_silver = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=@SI.1").json()["barData"]["priceBars"])
    df_silver["datetime"] = pd.to_datetime(df_silver['tradeTimeinMills'],unit='ms').dt.date
    df_silver.set_index("datetime",inplace = True)
    df_silver["close"] = df_silver["close"].astype(float)

    gold_silver_ratio = df_gold["close"]/df_silver["close"]
    
    df_bnd = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=BND").json()["barData"]["priceBars"])
    df_bnd["datetime"] = pd.to_datetime(df_bnd['tradeTimeinMills'],unit='ms').dt.date
    df_bnd.set_index("datetime",inplace = True)
    df_bnd["close"] = df_bnd["close"].astype(float)
    
    df_sphb = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=SPHB").json()["barData"]["priceBars"])
    df_sphb["datetime"] = pd.to_datetime(df_sphb['tradeTimeinMills'],unit='ms').dt.date
    df_sphb.set_index("datetime",inplace = True)
    df_sphb["close"] = df_sphb["close"].astype(float)
    
    df_splv = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=SPLV").json()["barData"]["priceBars"])
    df_splv["datetime"] = pd.to_datetime(df_splv['tradeTimeinMills'],unit='ms').dt.date
    df_splv.set_index("datetime",inplace = True)
    df_splv["close"] = df_splv["close"].astype(float)

    sphb_splv_ratio = df_sphb["close"]/df_splv["close"]
    
    df_HYG = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=HYG").json()["barData"]["priceBars"])
    df_HYG["datetime"] = pd.to_datetime(df_HYG['tradeTimeinMills'],unit='ms').dt.date
    df_HYG.set_index("datetime",inplace = True)
    df_HYG["close"] = df_HYG["close"].astype(float)
    
    df_fnda = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=FNDA").json()["barData"]["priceBars"])
    df_fnda["datetime"] = pd.to_datetime(df_fnda['tradeTimeinMills'],unit='ms').dt.date
    df_fnda.set_index("datetime",inplace = True)
    df_fnda["close"] = df_fnda["close"].astype(float)
    
    df_schx = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=SCHX").json()["barData"]["priceBars"])
    df_schx["datetime"] = pd.to_datetime(df_schx['tradeTimeinMills'],unit='ms').dt.date
    df_schx.set_index("datetime",inplace = True)
    df_schx["close"] = df_schx["close"].astype(float)

    fnda_schx_ratio = df_fnda["close"]/df_schx["close"]
    
    df_btc_usd = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{tf}.json?symbol=BTC.CB=").json()["barData"]["priceBars"])
    df_btc_usd["datetime"] = pd.to_datetime(df_btc_usd['tradeTimeinMills'],unit='ms').dt.date
    df_btc_usd.set_index("datetime",inplace = True)
    df_btc_usd["close"] = df_btc_usd["close"].astype(float)
    periods = 300
    roro = []
    if tf == "5Y":
        state = 1
    else:
        state = 0
    for i in range(periods,0,-1):
        temp = dict(
            date = (pd.to_datetime(df_spx.index).date)[-i],
            spx = roro_comp_get(df_spx["close"],i,state),
            audjpy = roro_comp_get(df_AUDJPY["close"],i,state),
            gold_silver = roro_comp_get(gold_silver_ratio,i,state,inverse = True),
            bnd = roro_comp_get(df_bnd["close"],i,state,inverse = True),
            sphb_splv = roro_comp_get(sphb_splv_ratio,i,state),
            hyg = roro_comp_get(df_HYG["close"],i,state),
            fnda_schx = roro_comp_get(fnda_schx_ratio,i,state),
            vix = vix_gradient(df_vix["close"].iloc[-i]),
            btc_usd = roro_comp_get(df_btc_usd["close"],i,state)
        )
        roro.append(temp)

    return roro
# Setting the page layout as wide
st.set_page_config(layout="wide")
def get_data_yields(symbol,lookback_period):
    global response_yields
    df = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/{lookback_period}.json?symbol={symbol}").json()["barData"]["priceBars"])
    df["datetime"] = pd.to_datetime(df['tradeTimeinMills'],unit='ms')
    df.set_index("datetime",inplace = True)
    response_yields[symbol] = df["close"].astype(float)
def get_recommendation(symbol,rsi_val,drop_frm_ath,dist_from_5_yr_low):
    global momentum_recommendations,cheap_recommendations
    df = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/1Y.json?symbol={symbol}").json()["barData"]["priceBars"])
    df_all = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/ALL.json?symbol={symbol}").json()["barData"]["priceBars"])
    df_all["high"] = df_all["high"].astype(float)
    df_all["low"] = df_all["low"].astype(float) 
    df["tradeTimeinMills"] = pd.to_datetime(df['tradeTimeinMills'],unit='ms')
    df_all["tradeTimeinMills"] = pd.to_datetime(df_all['tradeTimeinMills'],unit='ms')
    df["close"] = df["close"].astype(float)
    df["open"] = df["open"].astype(float)
    df.set_index("tradeTimeinMills",inplace = True)
    df_all.set_index("tradeTimeinMills",inplace = True)

    current_close = df["close"].iloc[-1]

    df["50DMA"] = df["close"].rolling(50).mean()
    df["100DMA"] = df["close"].rolling(100).mean()
    df["Vol_mon_avg"] = 5*df["volume"].rolling(252).mean()
    df["RSI"] = talib.RSI(df["close"])
    cond1 = current_close>df["50DMA"].iloc[-1]
    cond2 = current_close>df["100DMA"].iloc[-1]
    cond3 = df["volume"].rolling(5).sum().iloc[-1]>1.5*df["Vol_mon_avg"].iloc[-1]
    mom_score = int(cond1)+int(cond2)+int(cond3)

    ath = df_all["high"].max()
    distance_from_ath = round((ath-current_close)*100/ath,2)
    yr_2_ago_dt =  datetime.now() - relativedelta(years=2)
    yr_2_low = df_all["low"].loc[yr_2_ago_dt:].min()
    distance_frm_2yr_low = round((current_close - yr_2_low)*100/current_close,2)
    rsi = round(df["RSI"].iloc[-1],2)
    cond_1 = rsi<rsi_val
    cond_2 = distance_from_ath>=drop_frm_ath
    cond_3 = distance_frm_2yr_low<=dist_from_5_yr_low
    cheap_score = int(cond_1)+int(cond_2)+int(cond_3)
    
    if (cond1 or cond2) or cond3:
        momentum_recommendations[symbol] = {"LTP":current_close,"50DMA":df["50DMA"].iloc[-1],"100DMA":df["100DMA"].iloc[-1],"Vol>1.5avg":cond3,"score":mom_score,"sparkline": sparkline(data=df["close"])}
    if cond_1 or cond_2 or cond_3:
        cheap_recommendations[symbol] = {"LTP":current_close,"RSI":rsi,f'drop frm ATH':distance_from_ath,f"% away 2 yr low":distance_frm_2yr_low,"score":cheap_score,"sparkline": sparkline(data=df["close"])}
def sparkline(data, figsize=(4,0.25),**kwags):
    data = list(data)
    fig,ax = plt.subplots(1,1,figsize=figsize,**kwags)
    ax.plot(data)
    
    for k,v in ax.spines.items():
        v.set_visible(False)
    
    ax.set_xticks([])
    ax.set_yticks([])
    
    plt.plot(len(data)-1, data[len(data)-1], 'r.')
    
    ax.fill_between(range(len(data)), data, len(data)*[min(data)], alpha=0.1)
    
    img = BytesIO()
    plt.savefig(img, transparent=True, bbox_inches='tight')
    img.seek(0)
#     plt.show()
    plt.close()
    
    # return base64.b64encode(img.read()).decode("utf-8")
    return '<img src="data:image/png;base64,{}"/>'.format(base64.b64encode(img.read()))
def highlight_rec_momentum(s):
    arr = []
    arr.append('background-color: white')
    if s["50DMA"]<s["LTP"]:
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    if s["100DMA"]<s["LTP"]:
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    if s["Vol>1.5avg"]:       
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    arr.append('background-color: white')
    arr.append('background-color: white')
    return arr
def font_color(s):
    return ["color: black"]*len(s)
def highlight_rec_cheap(s,rsi_val,drop_frm_ath,dist_from_5_yr_low):
    arr = []
    arr.append('background-color: white')
    if s["RSI"]<rsi_val:
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    if s[f"drop frm ATH"]>drop_frm_ath:
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    if s[f"% away 2 yr low"]<dist_from_5_yr_low:       
        arr.append('background-color: #90EE90')
    else:
        arr.append('background-color: #FF7F7F')
    arr.append('background-color: white')
    arr.append('background-color: white')

    return arr
def get_etf_rets(symbol):
    global res_etf_ret, vol_etf_info,expense_ratios
    df = pd.DataFrame(requests.get(f"https://ts-api.cnbc.com/harmony/app/charts/1Y.json?symbol={symbol}").json()["barData"]["priceBars"])
    df["tradeTimeinMills"] = pd.to_datetime(df['tradeTimeinMills'],unit='ms')
    df["close"] = df["close"].astype(float)
    df["open"] = df["open"].astype(float)
    df.set_index("tradeTimeinMills",inplace = True)
    df_new = df.resample("W-Sun").agg({"close":"last","open":"first"})
    price_1_yr_ago = df.loc[df_new.index[-1] - relativedelta(years = 1):].iloc[0]["open"]
    price_1_mon_ago = df.loc[df.index[-1] - relativedelta(months=1):].iloc[0]["open"]
    close = df.iloc[-1]["close"]
    daily_ret = (close - df.iloc[-1]["open"].astype(float))*100/df.iloc[-1]["open"]
    w_1_ret = round((close - df_new.iloc[-1]["open"])*100/df_new.iloc[-1]["open"],2)
    w_2_ret = round((close - df_new.iloc[-2]["open"])*100/df_new.iloc[-2]["open"],2)
    y_1_ret = round((close - price_1_yr_ago)*100/price_1_yr_ago,2)
    m_1_ret = round((close - price_1_mon_ago)*100/price_1_mon_ago,2)
    temp = dict(symbol = symbol,day_ret = daily_ret, w1_ret = w_1_ret, w2_ret = w_2_ret, year_ret = y_1_ret,m1_ret = m_1_ret)
    res_etf_ret.append(temp)
    iv = (df["close"].pct_change()*100).iloc[-30:].std()*np.sqrt(252)
    vol_etf_info[symbol] = round(iv,2)

    r = requests.get(f"https://etfdb.com/etf/{symbol}/#etf-ticker-profile")
    soup = BeautifulSoup(r.content, 'html5lib')
    expense_ratio = soup.find("div",{"class":"ticker-assets"}).find_all("div")[3].text.split("\n")[-2]
    expense_ratios[symbol] = expense_ratio
    return 
def get_data(ticker, timeframe = 60*60 * 4):
    # Function to get OHLC data for a symbol from FTX api
    data = pd.DataFrame(json.loads(requests.get(f"https://ftx.com/api/markets/{ticker}/candles?resolution={timeframe}").text)["result"])
    return data
def in_squeeze(symbol,bb_mul, kc_mul,num_days,plot = False, timeframe = 60*60*4):
    # Function to check whether Keltner Channel and Bollinger bands squeeze is happening
    #     Get Data
    data = get_data(symbol,timeframe)
#     Calculate BB
    data["20sma"] = data["close"].rolling(window = 20).mean()
    data["stddev"] = data["close"].rolling(window = 20).std()
    data["lowerband"] = data["20sma"] - bb_mul*data["stddev"]
    data["upperband"] = data["20sma"] + bb_mul*data["stddev"]
    
#     Calculate KC
    data["TR"] = data["high"] - data["low"]
    data["ATR"] = data["TR"].rolling(window = 20).mean()
    data['upperKC'] = data["20sma"] + kc_mul*data["ATR"]
    data['lowerKC'] = data["20sma"] - kc_mul*data["ATR"]
    data["squeeze_on"] = np.where(
                        np.logical_and(data["lowerband"]>data["lowerKC"], data["upperband"]<data["upperKC"]),1,0)
    # Now if "num_days" days earlier BB were in KC but now, its not then that's breakout
    if data.iloc[-num_days]["squeeze_on"] and not data.iloc[-1]["squeeze_on"]:
        # If user wants to plot the candlestick then pass the plot = True
        if plot == True:
            # Template for plotting KC and BB and candlesticks
            candlestick  = go.Candlestick(x=data["startTime"],open=data["open"],high=data["high"],low=data["low"],close=data["close"], name = symbol)
            upperband = go.Scatter(x = data["startTime"], y = data["upperband"], name = "Upper BB", line = dict(color = "blue"))
            lowerband = go.Scatter(x = data["startTime"], y = data["lowerband"], name = "Lower BB", line = dict(color = "blue"))
            upperKC = go.Scatter(x = data["startTime"], y = data["upperKC"], name = "Upper KC", line = dict(color = "green"))
            lowerKC = go.Scatter(x = data["startTime"], y = data["lowerKC"], name = "Upper KC", line = dict(color = "green"))

            fig = go.Figure(data = [candlestick,upperband, lowerband, upperKC, lowerKC])
            # Slider in Xaxes for 1month, 6month, YTD and 1Y
            fig.update_xaxes(
                rangeslider_visible=True,
                rangeselector=dict(
                    buttons=list([
                        dict(count=1, label="1mon", step="month", stepmode="backward"),
                        dict(count=6, label="6mon", step="month", stepmode="backward"),
                        dict(count=1, label="YTD", step="year", stepmode="todate"),
                        dict(count=1, label="1y", step="year", stepmode="backward"),
                        dict(step="all")
                    ])
            ))
            fig.update_layout(yaxis=dict(autorange = True,fixedrange= False))
            st.plotly_chart(fig,use_container_width=True)
        else:
            print(f"{symbol}")
        
def plot_candlestick(data):
    # Function to plot candlestick out of a given OHLC dataframe
    fig = go.Figure(data=[go.Candlestick(x=data.index,
                                    open=data['open'],
                                    high=data['high'],
                                    low=data['low'],
                                    close=data['close'])])
    fig.update_xaxes(
        rangeslider_visible=True,
        rangeselector=dict(
            buttons=list([
                dict(count=1, label="1mon", step="month", stepmode="backward"),
                dict(count=6, label="6mon", step="month", stepmode="backward"),
                dict(count=1, label="YTD", step="year", stepmode="todate"),
                dict(count=1, label="1y", step="year", stepmode="backward"),
                dict(step="all")
            ])
    ))
    return fig

def get_key(dict_given, val):
    # Function to get the key of the required value from a dictionary
    for key, value in dict_given.items():
        if val == value:
            return key
def get_crypto_data_daily(symbol):
    data = get_data(symbol,timeframe = 60*60*24)
    data.to_csv(f"crypto_data/{symbol.replace('/','_')}.csv")
def get_cnbc_yields(symbol):
    # Get the symbol quote from yahoo finance, we are using Beautiful Soup for scraping
    
    URL = f"https://www.cnbc.com/quotes/US5Y"
    headers = {'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.61 Safari/537.36'}
    page = requests.get(URL, headers = headers)
    soup = BeautifulSoup(page.text, "html.parser")
    yields = ["US 2Y","US 5Y", "US 10Y","US 30Y"]
    if symbol in yields:
        elements = soup.find('div',{'class':'QuoteStrip-lastPriceStripContainer'})
        price = elements.find_all('span')[0].text.replace("%","")
    return float(price.replace(",",""))
def get_yahoo_finance_quote(symbol):
    # Get the symbol quote from yahoo finance, we are using Beautiful Soup for scraping
    URL = f"https://finance.yahoo.com/quote/{symbol}"
    headers = {'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.61 Safari/537.36'}
    page = requests.get(URL, headers = headers)
    soup = BeautifulSoup(page.text, "html.parser")
    price = soup.find('div',{'class':'D(ib) Mend(20px)'}).find_all('fin-streamer')[0].text
    return price.replace(",","")
def get_symbol_quote(ticker_tape, symbol,tck_tape, idx):
    # Get the symbol quote from yahoo finance, we are using Beautiful Soup for scraping
    URL = f"https://finance.yahoo.com/quote/{symbol}"
    headers = {'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.61 Safari/537.36'}
    page = requests.get(URL, headers = headers)
    soup = BeautifulSoup(page.text, "html.parser")
    price = soup.find('div',{'class':'D(ib) Mend(20px)'}).find_all('span')[0].text
    change = soup.find('div',{'class':'D(ib) Mend(20px)'}).find_all('span')[1].text
    change = change.split(" ")[1]
    change = change[1:-1]
    if ticker_tape == "US 10Y":
        # US 10Y treated differently because of some formating purposes
        change = float(change.replace("(","").replace(")","").replace("%",""))
        price_now = float(price)
        change = round(price_now - price_now/(1+change/100),2)
        change = str(change) + "%"
    tck_tape[idx].metric(label = ticker_tape ,value = price,delta = change)
    
# Set the twitter client and set the access token which comes from config.py
auth = tweepy.OAuthHandler(TWITTER_CONSUMER_KEY, TWITTER_CONSUMER_SECRET)
auth.set_access_token(TWITTER_ACCESS_TOKEN, TWITTER_ACCESS_TOKEN_SECRET)

api = tweepy.API(auth)

if "trade_api" not in st.session_state:
    # Now, if this api is not there in session_state then add it,
    # we are doing this because we don't want to let the info go once site is refreshed
    # And session_state is kind of memory used for caching required info
    # Set the trade_api variable
    st.session_state.trade_api = REST(API_KEY, SECRET_KEY, API_URL)
if "stocks" not in st.session_state:
    # If stocks are not present then from companies.csv file add them
    with open('datasets/companies.csv') as f:
        companies = f.read().splitlines()
        symbols = []
        for company in companies:
            symbols.append(company.split(",")[0])
        st.session_state.stocks = symbols
if "login" not in st.session_state:
    # If user has not logged in set login = False
    # This is done because once logged in we want the program to remember that it logged in
    # Hence setting the session_state
    st.session_state.login = False

if st.session_state.login == False:
    # If login is false then create a login window
    st.sidebar.write("Login")

    # Create sidebar login boxes
    user = st.sidebar.text_input(label = "Username")
    password = st.sidebar.text_input(label = "Password", type = "password")
    login_btn = st.sidebar.button("Login")
    if login_btn:
        # This is the users list which will help in login and password for both is 123
        if user in ["Chaitanya", "Sagar"] and password == "Wizards@123Trade#":
            st.sidebar.success(f"Logged In as {user}")
            st.session_state.login = True
            # Now if clicked on login wait for 1 second then reload the page using experimental_rerun
            time.sleep(1)
            st.experimental_rerun()
        else:
            # If incorrect credentials then pose an error
            st.sidebar.warning("Incorrect Username/Password")
else:
    # Get the crypto data from FTX api
    df = pd.DataFrame(json.loads(requests.get("https://ftx.com/api/markets").text)["result"])
    # Get the LTP and 24H %change for BTCUSDT, ETHUSDT
    st.session_state.index_btc = df[df["name"]=="BTC/USDT"]["last"].values[0]
    st.session_state.index_btc_pct_change = round(df[df["name"]=="BTC/USDT"]["change24h"].values[0],2)
    st.session_state.index_eth = df[df["name"]=="ETH/USDT"]["last"].values[0]
    st.session_state.index_eth_pct_change = round(df[df["name"]=="ETH/USDT"]["change24h"].values[0],2)
    st.session_state.roro = get_roro()
    roro_df = pd.DataFrame(st.session_state.roro)
    roro_df.set_index("date",inplace = True)
    roro_df["sum"] = roro_df.sum(axis = 1) / len(roro_df.columns)
    # Code below is for formatting purpose
    cols_ticker_tape_cryp = st.columns([1,1])
    cols_ticker_tape_cryp[0].metric(label = "BTC",value = st.session_state.index_btc, delta = f"{st.session_state.index_btc_pct_change}%")
    cols_ticker_tape_cryp[1].metric(label = "ETH", value = st.session_state.index_eth, delta = f"{st.session_state.index_eth_pct_change}%")
    fig = go.Figure(go.Indicator(
        mode = "number+delta",
        value = round(roro_df["sum"].iloc[-1]),
        domain = {'x': [0, 0.2], 'y': [0, 0.2]},
        delta = {'reference': round(roro_df["sum"].iloc[-2])},
        gauge = {'axis': {'range': [-100, 100]}},
        title = {'text': "RORO Indicator"}))
    fig.add_trace(go.Scatter(y = roro_df["sum"].values,x = roro_df.index,mode = "lines",fill='tozeroy'))
    st.plotly_chart(fig,use_container_width=True,use_container_height = True)
    cols_ticker_tape = st.columns(5)
    # Now, since fetching each symbol will take lots of time
    # So, we are running all the process in threads parally multi-processing
    thread_ticker = []
    for i in range(len  (symbol_mapping.keys())):
        # Get the symbol_mapping and run it in thread and display each of them
        ticker_tape = list(symbol_mapping.keys())[i]
        t_ticker = threading.Thread(target = get_symbol_quote, args = (ticker_tape, symbol_mapping[ticker_tape],cols_ticker_tape,i,))
        add_report_ctx(t_ticker)
        t_ticker.start()
        thread_ticker.append(t_ticker)
    for x in thread_ticker:
        # Wait for the threads to finish
        x.join()
    # This is a menu for various dashboard windows
    option = st.sidebar.selectbox("Which Dashboard?", 
    ("Watchlist","twitter","wallstreetbets","stocktwits","CryptoIndex","Chart","pattern", "MACRO",
    "Technical Scanner", "coinBaskets", "Breakout","ETFs", "Commodities","Report","Recommendations","RORO Components"))
    # If you want to momentarily hide your website from people then remove the below from commenting
    # Or you could do one more thing, setup a pseudo id-password, which log in you to only limited features
    # option = st.sidebar.selectbox("Which Dashboard?", 
    # ("twitter","coinBaskets"))
    # Set the option value as header
    st.header(option)
    if option == "Watchlist":
        # If watchlist is selected
        st.subheader("Watchlist")
        col1, col2 = st.columns([1,6.5])
        df_watchlist = pd.read_csv('watchlist.csv')
        with col1:
            # Now, create a form which will help you add symbol name, comments, etc
            with st.form(key = "symbol_add"):
                # df = pd.DataFrame(json.loads(requests.get("https://ftx.com/api/markets").text)["result"])
                # df = df[df["quoteCurrency"].isin(["USD","USDT"])]
                # symbols = df.name.values
                # asset_class = st.sidebar.selectbox("Asset Class",["Crypto","Fixed Income","Stocks","Index","Commodity"])
                asset_class = st.sidebar.selectbox("Asset Class",["Fixed Income","Stocks","Index","Commodity"])
                
                if asset_class == "Crypto":
                    symbol = st.selectbox("Symbol",symbols)
                    st.write("Currently working for crypto symbols from FTX")
                elif asset_class == "Fixed Income":
                    symbol = st.selectbox("Symbol",["HYG","LQD","US 2Y","US 5Y","US 10Y","US 30Y"])   
                elif asset_class == "Index":
                    symbol = st.selectbox("Symbol",["SPX","NASDAQ","NIFTY50","VIX"])             
                st.write("If an alert got triggered, but you want to add another, first delete it, then add")
                trigger = st.text_input("Trigger Price")
                view_type = st.selectbox("View Type",["Above","Below"])
                dma200 = st.selectbox("DMA alert",["Yes", "No"])
                dma200_view_type = st.selectbox("DMA200 View Type",["Above","Below"])
                alert_type = st.selectbox("Alert Type",["Macro","Individual"])
                comments = st.text_area("Comments")
                # Create buttons like add and delete symbols for use
                add_symbol = st.form_submit_button("Add")
                delete_symbol = st.form_submit_button("Remove")
                # Empty watchlist button to clear whole watchlist
                empty_checkbox = st.checkbox("Yes, I wish to empty the watchlist",value = False)
                empty_watchlist = st.form_submit_button("Empty Watchlist")
        with col2:
            # If empty_watchlist button is clicked then, clear the watchlist
            if empty_watchlist and empty_checkbox:
                st.info("Watchlist cleared")
                df_watchlist = pd.DataFrame(columns=["Symbol","Comments"])
                df_watchlist.to_csv('watchlist.csv',index=False)
            elif empty_watchlist:
                st.error("Could not clear the watchlist, please select the checkbox")
            elif add_symbol:
                if symbol not in df_watchlist["Symbol"].values:
                    # If symbol is added then add it to csv file, then update the csv file
                    st.info("Symbol added")
                    df_watchlist = df_watchlist.append(dict(Symbol = symbol, Trigger = float(trigger), alert_type = alert_type, view_type = view_type, status = "Pending",dma200 = dma200, dma200_view_type = dma200_view_type,dma_status = "Pending",Comments = comments), ignore_index=True)
                    df_watchlist.to_csv('watchlist.csv',index=False)
                else:
                    st.warning("Symbol Already Present, Please check")
            elif delete_symbol:
                if symbol not in df_watchlist["Symbol"].values:
                    st.warning("Symbol not present, please check...")
                else:
                    # Delte symbol and update the csv file
                    st.info("Symbol Deleted")
                    df_watchlist = df_watchlist[df_watchlist["Symbol"]!= symbol]
                    df_watchlist.to_csv('watchlist.csv',index=False)
            if len(df_watchlist)>0:
                # If there's something in watchlist then for cryptos fetch the LTP from FTX api
                # Getting LTP is currently for Crypto only due to free data unavailability for stocks/ETFs
                df_watchlist.set_index("Symbol",inplace = True)
                df_watchlist["LTP"] = 0
                def get_data_symbol(symbol_name_for_quote):
                    global df_watchlist
                    try:
                        if symbol_name_for_quote in ["HYG","LQD","SPX","NASDAQ","NIFTY50","VIX","US 2Y","US 5Y", "US 10Y","US 30Y"]:
                            data_symbol = get_cnbc_data(symbol_name_for_quote)
                        else:
                            data_symbol = json.loads(requests.get(f"https://ftx.com/api/markets/{symbol_name_for_quote}").text)["result"]["price"]
                        df_watchlist.loc[symbol_name_for_quote,"LTP"] = data_symbol
                    except:
                        pass
                threads_list = []
                for i in range(len(df_watchlist)):
                    symbol_name_for_quote = df_watchlist.index[i]
                    x = threading.Thread(target = get_data_symbol,args = (symbol_name_for_quote,))
                    x.start()
                    add_report_ctx(x)
                    threads_list.append(x)
                for thread in threads_list:
                    thread.join()
                df_watchlist["pct_away"] = np.round(100*np.abs(df_watchlist["Trigger"] - df_watchlist["LTP"])/df_watchlist["LTP"])
            # Show the watchlist
            st.dataframe(df_watchlist)
    elif option == "RORO Components":
        lkbck_perd = st.sidebar.selectbox("Lookback Period",["1Y","5Y"])
        st.sidebar.write("1Y will fetch daily data, 5Y --> weekly")
        st.sidebar.write("Number of lookback periods is set as 300")
        roro_run_btn = st.sidebar.button("Run")
        if roro_run_btn:
            roro_comp_df = pd.DataFrame(get_roro(lkbck_perd))
            roro_comp_df.set_index("date",inplace = True)
            roro_comp_df["roro"] = roro_comp_df.sum(axis = 1)/(len(roro_comp_df.columns))
            fig = make_subplots(rows=len(roro_comp_df.columns), shared_xaxes=True,vertical_spacing=0.01,subplot_titles=(roro_comp_df.columns))
            counter = 1
            for key in roro_comp_df.columns:
                if key!= "roro":
                    fig.add_scatter(x = roro_comp_df.index, y = roro_comp_df[key].values,mode = "lines",row = counter, col = 1, name = key)
                else:
                    fig.add_scatter(y = roro_comp_df["roro"].values,x = roro_comp_df.index,mode = "lines",fill='tozeroy',row = counter,col =1,name = key)
                counter+=1
            fig['layout'].update(height=2500, width=600, title='Subplots of components')
            st.plotly_chart(fig,use_container_width = True)
    elif option == "Recommendations":
        select_asset = st.sidebar.selectbox("Asset Class",["ETF"])
        rsi_val = st.sidebar.number_input("RSI Thresh",value = 50)
        drop_frm_ath = st.sidebar.number_input("Drop From ATH(%) Thresh",value = 50)
        dist_from_5_yr_low = st.sidebar.number_input(f"%away from 2 yr Low Thresh",value = 20)
        rec_run_btn = st.sidebar.button("Run")
        st.sidebar.write("**Momentum Rising Screening Conditions**")
        st.sidebar.write("LTP>50DMA")
        st.sidebar.write("LTP>100DMA")
        st.sidebar.write("5D Vol>1.5*(5D Vol. yearly_avg")
        st.sidebar.write("**Cheap Stocks Screening Conditions**")
        st.sidebar.write("RSI < RSI Thresh")
        st.sidebar.write("Drop from ATH(%) > Drop from ATH(%) Thresh")
        st.sidebar.write(f"% away from 2 yr Low < %away from 2 yr Low Thresh")
        if rec_run_btn:
            momentum_recommendations = {}
            cheap_recommendations = {}
            rec_threads = []
            if select_asset == "ETF":
                options_for_assets = all_etfs
            for asset in options_for_assets:
                x = threading.Thread(target=get_recommendation,args = (asset,rsi_val,drop_frm_ath,dist_from_5_yr_low,))
                x.start()
                rec_threads.append(x)
                # get_recommendation(asset)
            for rec in rec_threads:
                rec.join()
            # st.write(momentum_recommendations)
            momentum_recommendations = pd.DataFrame(momentum_recommendations).T
            
            momentum_recommendations.sort_values("score",ascending=False,inplace = True)
            cheap_recommendations = pd.DataFrame(cheap_recommendations).T
            cheap_recommendations.sort_values("score",ascending=False,inplace = True)

            st.header("Momentum Rising")
            st.dataframe(momentum_recommendations.style.apply(highlight_rec_momentum,axis = 1).apply(font_color))
            st.header("Cheap Stocks")
            st.dataframe(cheap_recommendations.style.apply(highlight_rec_cheap,args = (rsi_val,drop_frm_ath,dist_from_5_yr_low),axis = 1).apply(font_color))
    elif option == "MACRO":
        select_timeframe = st.sidebar.selectbox("Which Timeframe (finviz only)?",["Daily","Weekly","Monthly"])
        # select_lookback_period = st.sidebar.selectbox("LookbackPeriod for yields(yrs)",[1,2,3,4,5])
        select_lookback_period = st.sidebar.selectbox("LookbackPeriod for yields",["1D","1M","3M","6M","1Y","5Y"])
        timeframe_map = {"Daily":"d1","Weekly":"w1","Monthly":"mo"}
        tf = timeframe_map[select_timeframe]
        run_btn = st.sidebar.button("Run")
        if run_btn:
            col1, col2, col3 = st.columns([1,1,1])
            with col1:
                st.image(f"https://finviz.com/fut_image.ashx?es_{tf}_s.png")
            with col2:
                st.image(f"https://finviz.com/fut_image.ashx?vx_{tf}_s.png")
            with col3:
                st.image(f"https://finviz.com/fut_image.ashx?nq_{tf}_s.png")
            response_yields = {}
            thread_yields = []
            for symbol in ["US1Y","US2Y","US5Y","US10Y","US30Y"]:
                thread = threading.Thread(target = get_data_yields, args = (symbol,select_lookback_period,))
                thread.start()
                thread_yields.append(thread)
            for x in thread_yields:
                x.join()
            df_yields = pd.concat(response_yields,axis = 1)
            df_yields["2y/10y"] = df_yields["US10Y"]-df_yields["US2Y"]
            df_yields["10y/30y"] = df_yields["US30Y"]-df_yields["US10Y"]
            
            df_yields["2y/5y"] = df_yields["US5Y"]-df_yields["US2Y"]
            df_yields["1y/2y"] = df_yields["US2Y"]-df_yields["US1Y"]

            
            st.line_chart(df_yields[["US1Y","US2Y","US5Y","US10Y","US30Y"]])
            col1, col2 = st.columns([1,1])
            with col1:
                st.line_chart(df_yields["1y/2y"])
                st.line_chart(df_yields["2y/5y"])
            with col2:
                st.line_chart(df_yields["2y/10y"])
                st.line_chart(df_yields["10y/30y"])
            # today = datetime.now().strftime("%Y-%m-%d")
            # past = (datetime.now() - timedelta(days=365*select_lookback_period)).strftime("%Y-%m-%d")
            # # https://fred.stlouisfed.org/graph/?id=DGS10,DGS5,DGS30,DGS3MO,DGS1,DGS2,
            # url = f"""https://fred.stlouisfed.org/graph/fredgraph.png?dwnld=0&hires=1&type=image/png&
            # bgcolor=%23e1e9f0&chart_type=line&drp=0&fo=open%20sans&graph_bgcolor=%23ffffff&height=450&
            # mode=fred&recession_bars=on&txtcolor=%23444444&ts=12&tts=12&width=1168&nt=0&thu=0&trc=0&
            # show_legend=yes&show_axis_titles=yes&show_tooltip=yes&id=DGS10,DGS5,DGS30,DGS3MO,DGS1,DGS2&
            # scale=left,left,left,left,left,left&cosd={past},{past},{past},{past},{past},{past}&
            # coed={today},{today},{today},{today},{today},{today}&
            # line_color=%234572a7,%23aa4643,%2389a54e,%2380699b,%233d96ae,%23db843d&
            # link_values=false,false,false,false,false,false&line_style=solid,solid,solid,solid,solid,solid&
            # mark_type=none,none,none,none,none,none&mw=3,3,3,3,3,3&lw=2,2,2,2,2,2&
            # ost=-99999,-99999,-99999,-99999,-99999,-99999&oet=99999,99999,99999,99999,99999,99999&
            # mma=0,0,0,0,0,0&fml=a,a,a,a,a,a&fq=Daily,Daily,Daily,Daily,Daily,Daily&
            # fam=avg,avg,avg,avg,avg,avg&fgst=lin,lin,lin,lin,lin,lin&
            # fgsnd=2020-02-01,2020-02-01,2020-02-01,2020-02-01,2020-02-01,2020-02-01&line_index=1,2,3,4,5,6&
            # transformation=lin,lin,lin,lin,lin,lin&
            # vintage_date={today},{today},{today},{today},{today},{today}&
            # revision_date={today},{today},{today},{today},{today},{today}&
            # nd=1962-01-02,1962-01-02,1977-02-15,1981-09-01,1962-01-02,1976-06-01
            # """
            # url = url.replace("\n","")
            # st.image(url,width = 800)


    elif option == "twitter":
        # If twitter is selected
        today = datetime.today()
        # Get the local timezone, this is important because then it works on a different timezone
        to_zone = tz.tzlocal()
        # Multibox for selecting multiple users
        who = st.sidebar.multiselect("Choose person",tuple(TWITTER_USERNAMES))
        twitter_run_btn = st.sidebar.button("Run")
        if twitter_run_btn:
            # if twitter run button is clicked then all those people selected are called
            if "SELECT ALL" in who:
                users_list = TWITTER_USERNAMES[1:]
            else:
                users_list = who
                
            for username in users_list:
                # For a given username fetch the tweets, its username, image
                user = api.get_user(screen_name = username)
                tweets = api.user_timeline(screen_name = username, count = 100, tweet_mode = "extended")
                st.subheader(username)
                st.image(user.profile_image_url)
                for tweet in tweets:
                    # In all his tweets, bring those to local timezone
                    tweet_date = ((tweet.created_at).astimezone(to_zone)).replace(tzinfo = None)
                    #Now, we don't want tweets older than 3 days
                    delta = (today - tweet_date).days
                    if delta>3:
                        continue
                    # For the following user names certain modification is done to get the tweets
                    if username in ["@chartmojo","@MacroCharts"]:
                        if tweet.in_reply_to_screen_name== None:
                            st.subheader(tweet._json["created_at"])
                            st.write(tweet.full_text)
                            try:
                                for j in tweet.extended_entities["media"]:
                                    st.image(j["media_url_https"], width=600)
                            except:
                                pass
                    else:
                        if tweet.in_reply_to_screen_name== None and len(tweet.entities["symbols"])>0:
                            symbols = []
                            for i in range(len(tweet.entities["symbols"])):
                                symbols.append(tweet.entities["symbols"][i]["text"])
                            st.subheader(" ".join(symbols))
                            st.subheader(tweet._json["created_at"])
                            st.subheader(tweet.full_text)
                            try:
                                for j in tweet.extended_entities["media"]:
                                    st.image(j["media_url_https"], width = 600)
                            except:
                                pass
                            for symbol in symbols:
                                st.image(f"https://finviz.com/chart.ashx?t={symbol}&ta=1", width=600)
    # elif option == "US Sectors":
    #     st.sidebar.write("Source - TradingView")
    #     select_sector = st.sidebar.selectbox("Select Sector", options = us_sectors)
    #     select_btn = st.sidebar.button("Run")
    #     url = "https://in.tradingview.com/markets/stocks-usa/sectorandindustry-sector/"
    #     r = requests.get(url)
    #     soup = BeautifulSoup(r.content, 'html5lib') # If this line causes an error, run 'pip install html5lib' or install html5lib
    #     req = soup.find_all('tr',attrs = {'class':'tv-data-table__row tv-data-table__stroke tv-screener-table__result-row'})
    #     sectors = []
    #     pct_change = []
    #     for i in req:    
    #         arr = []
    #         for sector in i.find_all("td"):
    #             arr.append(sector.text)
    #         arr[0]  = arr[0].split("\n")[3].split("\t")[0]
    #         sectors.append(arr[0])
    #         pct_change.append(float(arr[3][:-1]))
    #     fig = go.Figure([go.Bar(x=sectors, y=pct_change,
    #                     marker = dict(color = ['rgba(63, 195, 128, 1)' if value>0 else 'rgba(219, 10, 91, 1)' for value in pct_change], 
    #                             line = dict(color='rgb(0,0,0)',width=1.5)))])
    #     st.plotly_chart(fig,use_container_width=True)
    #     if select_btn:
    #         val = "-".join(select_sector.lower().split(" "))
    #         url = f"https://in.tradingview.com/markets/stocks-usa/sectorandindustry-sector/{val}/"
    #         r = requests.get(url)
    #         soup = BeautifulSoup(r.content, 'html5lib') # If this line causes an error, run 'pip install html5lib' or install html5lib
    #         req = soup.find_all('tr',attrs = {'class':'tv-data-table__row tv-data-table__stroke tv-screener-table__result-row'})
    #         ticker = []
    #         pct_change_ticker = []
    #         company = []
    #         for i in req:    
    #             arr = []
    #             for sector in i.find_all("td"):
    #                 arr.append(sector.text)
    #             arr[0]  = arr[0].split("\n")[4]
    #             # result.append(dict(ticker = arr[0],pct_change = float(arr[2][:-1]),vol = arr[5],mkt_cap = arr[6]))
    #             # ticker.append(arr[0]) 
    #             company.append(us_stocks_mapping[arr[0]])
    #             pct_change_ticker.append(float(arr[2][:-1]))
    #         layout = go.Layout(
    #             xaxis = go.XAxis(
    #                 title = "Stocks",
    #                 showticklabels = False
    #             )
    #         )
    #         fig = go.Figure([go.Bar(x=company, y=pct_change_ticker,
    #                     marker = dict(color = ['rgba(63, 195, 128, 1)' if value>0 else 'rgba(219, 10, 91, 1)' for value in pct_change_ticker], 
    #                             line = dict(color='rgb(0,0,0)',width=1.5)))],layout = layout)
    #         st.write(len(company))
    #         st.plotly_chart(fig,use_container_width=True)
        
    #     st_autorefresh(interval=120000, limit=10000, key="US sectors refresh")
    elif option == "Commodities":
        select_timeframe = st.sidebar.selectbox("Which Timeframe?",["Daily","Weekly","Monthly"])
        timeframe_map = {"Daily":"d1","Weekly":"w1","Monthly":"mo"}
        tf = timeframe_map[select_timeframe]
        col0,col1,col2 = st.columns([1,1,1])
        count = 0
        for key in commodity_mapping:
            # https://finviz.com/futures_charts.ashx?t=YM&p=d1
            # Get the mapping for commodities, which is there in config.py
            keyword_comm = commodity_mapping[key]
            num = count%3
            if num ==0:
                with col0:
                    st.image(f"https://finviz.com/fut_image.ashx?{keyword_comm}_{tf}_s.png")
            elif num == 1:
                with col1:
                    st.image(f"https://finviz.com/fut_image.ashx?{keyword_comm}_{tf}_s.png")
            else:
                with col2:
                    st.image(f"https://finviz.com/fut_image.ashx?{keyword_comm}_{tf}_s.png")
            count = count+1
    elif option == "CryptoIndex":
        # If cryptoIndex tab is selected
        # Select which crypto index, then timeframe and accordingly fetch the data
        type_index = st.sidebar.selectbox("Which?",["Major","Minor","Shit"])
        select_index_timeframe = st.sidebar.selectbox("Timeframe",["15s","1m","5m","15m","1h","4h","1d"])
        if select_index_timeframe[-1] == "s":
            timeframe = int(select_index_timeframe[:-1])
        elif select_index_timeframe[-1] == "m":
            timeframe = int(select_index_timeframe[:-1]) * 60
        elif select_index_timeframe[-1] == "h":
            timeframe = int(select_index_timeframe[:-1])*60*60
        elif select_index_timeframe[-1] == "d":
            timeframe = int(select_index_timeframe[:-1])* 60*60*24
        if type_index == "Major":
            # Our crypto index is of 0.5BTC + 0.5 ETH
            st.write("0.5BTC + 0.5ETH")
            # Fetch data and create crypto index in same proportion
            data_btc = pd.DataFrame(json.loads(requests.get(f"https://ftx.com/api/markets/BTC/USDT/candles?resolution={timeframe}").text)["result"])
            data_eth = pd.DataFrame(json.loads(requests.get(f"https://ftx.com/api/markets/ETH/USDT/candles?resolution={timeframe}").text)["result"])
            data_btc.set_index("startTime",inplace = True)
            data_eth.set_index("startTime",inplace = True)
            # Note crypto index are normalized
            data = (data_btc["close"]*0.5/data_btc["close"][0] + data_eth["close"]*0.5/data_eth["close"][0])*100
            fig = go.Figure()
            fig.add_trace(go.Scatter(x=data.index, y=data.values,
                                            mode='lines',
                                            name=type_index))
            fig.update_xaxes(
                rangeslider_visible=True,
                nticks = 20,
                spikemode = "toaxis",
                rangeselector=dict(
                    buttons=list([
                        dict(count=1, label="1mon", step="month", stepmode="backward"),
                        dict(count=6, label="6mon", step="month", stepmode="backward"),
                        dict(count=1, label="YTD", step="year", stepmode="todate"),
                        dict(count=1, label="1y", step="year", stepmode="backward"),
                        dict(step="all")
                    ])
                )
            )
            fig.update_layout(
                xaxis_tickformat = '%Y-%m-%d',
                height = 600,
                width  = 900,
                hovermode = "x"
            )
            st.plotly_chart(fig)

    elif option == "Report":

        df = pd.DataFrame(json.loads(requests.get("https://ftx.com/api/markets").text)["result"])
        df = df[df["quoteCurrency"].isin(["USD","USDT"])]
        symbols = df.name.values
        
        to_analyze_symbol = st.sidebar.multiselect("Which symbol",symbols)
        analysis_date = st.sidebar.selectbox("Which date for Analysis",["Current","Yesterday"])
        if analysis_date == "Yesterday":
            locn = -2
        elif analysis_date == "Current":
            locn = -1
        fetch_data_btn = st.sidebar.button("Fetch Data")
        
        if fetch_data_btn:
            threads = []
            for symbol in symbols:
                t = threading.Thread(target = get_crypto_data_daily, args = (symbol,))
                t.start()
                add_report_ctx(t)

                threads.append(t)
            for x in threads:
                x.join()
            res = {}
            for req_symbol in to_analyze_symbol:
                req_symbol_file = req_symbol.replace("/","_")
                df_req = pd.read_csv(f"crypto_data/{req_symbol_file}.csv",index_col = 0, parse_dates = True)
                ans = {}
                for period in [20,50,100,200]:
                    df_req[f"MA{period}"] = talib.SMA(df_req["close"],period)
                    ans[f"% from MA{period}"] = round((df_req["close"].iloc[locn] - df_req[f"MA{period}"].iloc[locn])*100/df_req[f"MA{period}"].iloc[locn],2)
                df_req["RSI14"] = talib.RSI(df_req["close"])
                ans["RSI"] = round(df_req["RSI14"].iloc[locn],2)
                ans["LTP"] = df_req["close"].iloc[-1]
                res[req_symbol] = ans
            result = pd.DataFrame(res).T
            st.dataframe(result)
                    
    elif option == "Breakout":
        # If breakout tab is selected
        # Get the crypto symbols
        #Create buttons and boxes for selecting timeframe, BB number, KC multiplier
        df = pd.DataFrame(json.loads(requests.get("https://ftx.com/api/markets").text)["result"])
        df = df[df["quoteCurrency"].isin(["USD","USDT"])]
        symbols = df.name.values
        select_consol_tf = st.sidebar.selectbox("Timeframe",["15s","1m","5m","15m","1h","4h","1d"])
        if select_consol_tf[-1] == "s":
            timeframe = int(select_consol_tf[:-1])
        elif select_consol_tf[-1] == "m":
            timeframe = int(select_consol_tf[:-1]) * 60
        elif select_consol_tf[-1] == "h":
            timeframe = int(select_consol_tf[:-1])*60*60
        elif select_consol_tf[-1] == "d":
            timeframe = int(select_consol_tf[:-1])* 60*60*24
        select_BB_mul = st.sidebar.text_input("Bollinger Band multiplier",value = "2")
        select_KC_mul = st.sidebar.text_input("KC multiplier", value = "1.5")
        select_num_days = st.sidebar.text_input("Consolidating before how many periods?",3)
        num_days = int(select_num_days)
        bb_mul = float(select_BB_mul)
        kc_mul = float(select_KC_mul)
        consol_run_btn = st.sidebar.button("Run")
        if consol_run_btn:
            # If user clicks on run button
            # Again running in threads.
            threads = []
            for symbol in symbols:
                t = threading.Thread(target = in_squeeze, args = (symbol,bb_mul, kc_mul, num_days, True,60*60*24))
                add_report_ctx(t)
                t.start()
                threads.append(t)
            for x in threads:
                x.join()
            st.sidebar.write("Task Complete")
    elif option == "ETFs":
        # If ETFs are selected
        # Button to select ETF
        selectETF = st.sidebar.selectbox("Select ETF class",etf.keys())
        today = datetime.now() - timedelta(days =80)
        st.subheader(selectETF)
        # Note Country ETFs are dealt differently as they have country names as well
        if selectETF!= "Country":
            # If ETF is not Country then its very easy just from ETF variable we can fetch it
            etf_names = etf[selectETF].keys()
        else:
            # Select Markets, Country
            selectMarket = st.sidebar.selectbox("Which Market?",etf[selectETF].keys())
            country_list = list(etf[selectETF][selectMarket].keys())
            etf_list = list(etf[selectETF][selectMarket].values())
            etf_names = []
            country_names = []
            # In a for loop fetch all the ETFs for selected entries
            for i in range(len(etf_list)):
                if type(etf_list[i])==str:
                    etf_names.append(etf_list[i])
                    country_names.append(country_list[i])
                else:
                    for sub_etf_name in etf_list[i]:
                        etf_names.append(sub_etf_name)
                        country_names.append(country_list[i])
        select_timeframe = st.sidebar.selectbox("Which Timeframe?",["Daily","Weekly","Monthly"])
        timeframe_map = {"Daily":"d","Weekly":"w","Monthly":"m"}
        etf_run_btn = st.sidebar.button("Run")
        if etf_run_btn:
            tf = timeframe_map[select_timeframe]
            # If After giving all the entries, run button is clicked get the charts for all the ETFs
            count = 0
            res_etf_ret = []
            thread_det = []
            vol_etf_info = {}
            expense_ratios = {}
            st.info("Loading..... Please Have Patience")
            for n in etf_names:
                thread = threading.Thread(target = get_etf_rets, args = (n,))
                thread.start()
                thread_det.append(thread)
            for x in thread_det:
                x.join()
            st.success("Finished Loading")
            
            etf_rets_df = pd.DataFrame(res_etf_ret)
            etf_rets_df.set_index("symbol",inplace=True)
            col1,col2,col3,col4,col5 = st.columns([1,1,1,1,1])
            with col1:
                # st.write("")
                st.markdown("<h5 style='text-align: center; color: red;'>Daily Returns</h5>", unsafe_allow_html=True)
                st.dataframe(etf_rets_df["day_ret"].sort_values(ascending=False))
            with col2:
                st.markdown("<h5 style='text-align: center; color: red;'>1 Week Returns</h5>", unsafe_allow_html=True)
                
                st.dataframe(etf_rets_df["w1_ret"].sort_values(ascending=False))
            with col3:
                st.markdown("<h5 style='text-align: center; color: red;'>2 Week Returns</h5>", unsafe_allow_html=True)
                st.dataframe(etf_rets_df["w2_ret"].sort_values(ascending=False))
            with col4:
                st.markdown("<h5 style='text-align: center; color: red;'>1 Month Returns</h5>", unsafe_allow_html=True)
                st.dataframe(etf_rets_df["m1_ret"].sort_values(ascending=False))
            with col5:
                st.markdown("<h5 style='text-align: center; color: red;'>1 Year Returns</h5>", unsafe_allow_html=True)
                st.dataframe(etf_rets_df["year_ret"].sort_values(ascending=False))
            
            st.dataframe(etf_rets_df)
            cols1,cols2 = st.columns([1,1])
            if selectETF!= "Country":
                for etf_name in etf_names:
                    num = count%2
                    if num == 0:
                        with cols1:
                            try:
                                st.write(f"{etf[selectETF][etf_name]}, IV = {vol_etf_info[etf_name]}, ER = {expense_ratios[etf_name]}")
                            except:
                                pass
                            if tf == "d":
                                st.image(f"https://finviz.com/chart.ashx?t={etf_name}&ta=1&p={tf}")
                            else:
                                st.image(f"https://finviz.com/chart.ashx?t={etf_name}&p={tf}")
                    else:
                        with cols2:
                            try:
                                st.write(f"{etf[selectETF][etf_name]}, IV = {vol_etf_info[etf_name]}, ER = {expense_ratios[etf_name]}")
                            except:
                                pass
                            if tf == "d":
                                st.image(f"https://finviz.com/chart.ashx?t={etf_name}&ta=1&p={tf}")
                            else:
                                st.image(f"https://finviz.com/chart.ashx?t={etf_name}&p={tf}")
                    count = count + 1
            else:
                for i in range(len(etf_names)):
                    try:
                        st.write(f"{country_names[i]}, IV = {vol_etf_info[etf_names[i]]}, ER = {expense_ratios[etf_names[i]]}")
                    except:
                        pass
                    if tf == "d":
                        st.image(f"https://finviz.com/chart.ashx?t={etf_names[i]}&ta=1&p={tf}")
                    else:
                        st.image(f"https://finviz.com/chart.ashx?t={etf_names[i]}&p={tf}")    
    elif option == "coinBaskets":
        # If coinBaskets is selected
        # Note, Mudrex was our reference here
        baskets = st.sidebar.multiselect(label = "Baskets",options=names, default =names[0])
        run_basket = st.sidebar.button("Run")
        select_crypto_timeframe = st.sidebar.selectbox("Crypto Timeframe", options =
                                ["1d","1m","3m","5m","15m","30m","1h","2h","4h","6h","8h","12h","3d","1w","1M"])
        check_symbol = st.sidebar.text_input("Symbol check")
        interval = select_crypto_timeframe
        # Once all the inputs are given, and if any of the basket is chosen
        if check_symbol != "":
            for bkt in names:
                # For each basket selected, get its components, fetch its price from Binance api
                #Then plot all components in a single chart 
                #Here, you are giving in the symbol name and program is finding whether that symbol is there 
                # in any of the basket or not
                if check_symbol in eval(bkt)["components"]:
                    fig_check = go.Figure()
                    st.write(bkt.upper())
                    cols = st.columns(len(eval(bkt)["components"]))
                    for i in eval(bkt)["components"]:
                        ticker = f'{i.upper()}USDT'
                        req_params = dict(symbol = ticker, interval = interval)
                        url = "https://api.binance.com/api/v3/klines"
                        data = pd.DataFrame(json.loads(requests.get(url,params = req_params).text))
                        data = data.iloc[:,0:5]
                        data.columns = ['datetime', 'open','high','low', 'close']
                        data.index = [datetime.fromtimestamp(x/1000) for x in data.datetime]
                        data["close"] = data["close"].astype(float)
                        df = (data["close"].pct_change() + 1).cumprod()
                        fig_check.add_trace(go.Scatter(x=df.index, y=df.values,
                                            mode='lines',
                                            name=i))
                        fig_check.update_xaxes(
                        rangeslider_visible=True,
                        nticks = 20,
                        spikemode = "toaxis",
                        rangeselector=dict(
                            buttons=list([
                                dict(count=1, label="1mon", step="month", stepmode="backward"),
                                dict(count=6, label="6mon", step="month", stepmode="backward"),
                                dict(count=1, label="YTD", step="year", stepmode="todate"),
                                dict(count=1, label="1y", step="year", stepmode="backward"),
                                dict(step="all")
                            ])
                        )
                        )
                        fig_check.update_layout(
                            xaxis_tickformat = '%Y-%m-%d',
                            height = 600,
                            width  = 900,
                            hovermode = "x"
                        )
                        fig_check.update_traces(
                            hovertemplate="<br>".join([
                                "Price: %{y}"
                                ]))
                    st.plotly_chart(fig_check)
        if run_basket:
            for basket in baskets:
                # Here, you are choosing the baskets and program is plotting the baskets
                st.write(basket.upper())
                st.table(eval(basket))
                # Create traces
                fig = go.Figure()
                for i in eval(basket)["components"]:
                    ticker = f'{i.upper()}USDT'
                    interval = select_crypto_timeframe
                    req_params = dict(symbol = ticker, interval = interval)
                    url = "https://api.binance.com/api/v3/klines"
                    data = pd.DataFrame(json.loads(requests.get(url,params = req_params).text))
                    data = data.iloc[:,0:5]
                    data.columns = ['datetime', 'open','high','low', 'close']
                    data.index = [datetime.fromtimestamp(x/1000) for x in data.datetime]
                    data["close"] = data["close"].astype(float)
                    df = (data["close"].pct_change() + 1).cumprod()
                    fig.add_trace(go.Scatter(x=df.index, y=df.values,
                                        mode='lines',
                                        name=i))
                    fig.update_xaxes(
                    rangeslider_visible=True,
                    nticks = 20,
                    spikemode = "toaxis",
                    rangeselector=dict(
                        buttons=list([
                            dict(count=1, label="1mon", step="month", stepmode="backward"),
                            dict(count=6, label="6mon", step="month", stepmode="backward"),
                            dict(count=1, label="YTD", step="year", stepmode="todate"),
                            dict(count=1, label="1y", step="year", stepmode="backward"),
                            dict(step="all")
                        ])
                    )
                    )
                    fig.update_layout(
                        xaxis_tickformat = '%Y-%m-%d',
                        height = 600,
                        width  = 900,
                        hovermode = "x"
                    )
                    fig.update_traces(
                        hovertemplate="<br>".join([
                            "Price: %{y}"
                            ]))
                st.plotly_chart(fig)
    elif option == "Chart":
        # Charting platform
        # Get user inputs
        symbols = st_tags_sidebar(label = "Choose the tickers", 
        text = 'Press enter to add more', maxtags = 100)
        select_stock_timeframe = st.sidebar.selectbox("Stock Timeframe",options=["1Min", "5Min", "15Min", "day"])
        select_crypto_timeframe = st.sidebar.selectbox("Crypto Timeframe", options =
                                ["1d","1m","3m","5m","15m","30m","1h","2h","4h","6h","8h","12h","3d","1w","1M"])
        select_periods = st.sidebar.text_input("Number of Days(Stock)",value = "30")
        today = datetime.now() - timedelta(int(select_periods))
        if len(symbols)>0:
            # If symbols are selected 
            for symbol in symbols:
                # Run over all symbol in for loop and if they are crypto, treat them differently
                #And if they are others, treat them differently
                if f'{symbol.upper()}' in crypto_symbols:
                    # If crypto, get the data from Binance api
                    ticker = f'{symbol.upper()}USDT'
                    interval = select_crypto_timeframe
                    req_params = dict(symbol = ticker, interval = interval)
                    url = "https://api.binance.com/api/v3/klines"
                    st.subheader(ticker)
                    data = pd.DataFrame(json.loads(requests.get(url,params = req_params).text))
                    data = data.iloc[:,0:5]
                    data.columns = ['datetime', 'open','high','low', 'close']
                    data.index = [datetime.fromtimestamp(x/1000) for x in data.datetime]
                    data.drop("datetime",axis = 1, inplace = True)
                else:
                    # If stocks then get it from trade_api
                    # Remember!!!--> trade_api was initiated in the session_state in the starting of code
                    data = st.session_state.trade_api.get_barset(symbol.upper(), select_stock_timeframe, start =today.strftime("%Y-%m-%d")).df
                    data = data[symbol.upper()]
                    st.subheader(symbol.upper())
                fig = plot_candlestick(data)
                st.plotly_chart(fig, use_container_width=False)

    elif option == "stocktwits":
        # If stocktwits is selected
        # This is almost similar to Larry's video, so you can reference that as well
        
        # Display Trending stocks based on watchlist count
        st.header("Most Trending Symbols")
        most_trending_syms = get_stocktwits_data(req = "https://api.stocktwits.com/api/2/charts/ts",
                            code = "ts", label = "Trending Score")
        st.dataframe(most_trending_syms)

        st.header("Most messages in last 24 hrs")
        most_active_syms = get_stocktwits_data(req = "https://api.stocktwits.com/api/2/charts/m_day",
                            code = "m_day", label = "#messages")
        st.dataframe(most_active_syms)
        
        st.header("Top New Watchers added in last 24 hrs")
        most_active_syms = get_stocktwits_data(req = "https://api.stocktwits.com/api/2/charts/wl_ct_day",
                            code = "wl_ct_day", label = "Count")
        st.dataframe(most_active_syms)

        # For a given symbol, use request module to hit the stocktwits api and get the required info
        
        

        symbol = st.sidebar.text_input("Symbol", value = "AAPL", max_chars = 5)
        r = requests.get(f"https://api.stocktwits.com/api/2/streams/symbol/{symbol}.json")
        data = r.json()
        for message in data["messages"]:
            st.image(message["user"]["avatar_url"])
            st.write(message['user']["username"])
            st.write(message["created_at"]) 
            st.write(message["body"])
        
        st.sidebar.write("Update time for -->")
        st.sidebar.write("Top Watchlist Counts : 5mins")
        st.sidebar.write("Most Messages : 1hr")

    elif option == "Technical Scanner":
        # If technical scanner is selected
        # Then choose the pattern, timeframe, etc
        pattern = st.sidebar.selectbox("Which Pattern?", tuple(patterns.values()))
        keywords = st_tags_sidebar(label = "Choose the tickers", 
        text = 'Press enter to add more', maxtags = 100)
        select_timeframe = st.sidebar.selectbox("Timeframe",options=["1Min", "5Min", "15Min", "day"])
        run_btn = st.sidebar.button("Run")
        # Get the mapping for patterns from pattern.py file
        pattern_code = get_key(patterns, pattern)
        pattern_function = getattr(talib,pattern_code)
        if run_btn:
            # If clicked on run button
            # Then for each symbol, check the output
            keyword = [x.upper() for x in keywords]
            data = st.session_state.trade_api.get_barset(keyword, select_timeframe, limit = 100).df
            for symbol in keyword:
                try:
                    # st.write(data)
                    result = pattern_function(data[symbol]["open"],data[symbol]["high"],data[symbol]["low"],data[symbol]["close"])
                    last = result.tail(1).values[0]
                    # For some of the indicators value <0 is Bearish and value>0 is Bullish
                    # But for others there would be a different logic, you shall handle it differently
                    if last>0:
                        st.write(f"Bullish {symbol}")
                    elif last<0:
                        st.write(f"Bearish {symbol}")
                except:
                    pass