screener_csv.py

"""
Note: This is a script used as example for the class:
- For computationally efficient script USE screener.py on
this repository.

Which topics we will work here:
1. System process (make folder - delete folder with data)
2. Data process (gather - cleasing - store in .csv - read .csv)
3. Dates introduction (TimeSeries)
4. List creation and append values
5. Function creations (Technical screener)
6. Conditional statments (screening process)
7. Usage of some librarires (investpy / yfinance - pandas (essentials) - TA)
8. Try/Except and introduction to errors.

This screener the skeleton for a basic market screener. It will help you with:

1. Market Technical Screening
2. Market Alerts Signals
3. Further Analysis you might require

Main difference is on this script you will NOT BE STORING data on your local machine. With it
pros and cons. Because it doesn't store data is computationally more efficient and fast. 

For the example, the output of the screener is a prompt print with 
the list of tickers stored on the list variables. You can use this lists for further analyisis or
add more.

Example on the script:

country ='Argentina' <- Change this country for the required one
days_back = 120 <- Data gathering from this day. Will impact on the indicators (200SMA won't work on 120 days of data ;-) )
stocks = investpy.get_stocks_overview(country, n_results=1000) <- n_results=1000 For wider markets go for bigger results.

"""



import pandas as pd
import os
import shutil
import investpy
import time
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')


# 1. Creamos la carpeta donde se almacenan temporalmente los .csv
# 1. We will create the path were our temporary data will be stored in .csv
if not os.path.exists('data'):
    os.mkdir('data')

# 2. Para este ejemplo, leeremos los activos de la propia base de datos.
# 2. On this particular script, we will read our data from the database. 

country ='Argentina'
days_back = 120
today = datetime.now()
start = today -timedelta(days_back)
today = datetime.strftime(today, '%d/%m/%Y')
start = datetime.strftime(start, '%d/%m/%Y')
stocks = investpy.get_stocks_overview(country, n_results=1000)
stocks = stocks.drop_duplicates(subset='symbol')

# 3. Seleccionamos la ruta en la que estará nuestro proyecto con el final en \data puesto que ahí se almacenará la información.
# 3. Select the path in where our project will be stored. Remember end on \data puesto que ahí se almacenará la información.
path = (r"data")

# 4. Manejo de fechas. Recordad modificar days_back si necesitamos indicadores con más longitud de datos (mm200, etc.)
# 4. Date handling. Remember to modify days_back in case we need indicators with longer data (sma200, etc.)

today = datetime.now()
start = today -timedelta(days_back)
today = datetime.strftime(today, '%d/%m/%Y')
start = datetime.strftime(start, '%d/%m/%Y')

# 5. Proceso de solicitud de datos a través de la librería investpy para el documento especies.csv columna (Ticker)
# 5. Data gathering process through investpy library for especies.csv on (Ticker) column

count = 0
for ticker in stocks['symbol']:
    try:
        count += 1
        df = investpy.get_stock_historical_data(stock=ticker,country=country,from_date=f'{start}', to_date=f'{today}')
        df= df.rename(columns={"Close": "Adj Close"})
        # print(f'Analyzing {count}.....{ticker}')
        # print(df.info())  <== To see what you're getting
        df.to_csv(fr'data/{ticker}.csv')
        time.sleep(0.25)
    except Exception as e:
        print(e)
        print(f'No data on {ticker}')


# 6. A continuación definiremos algún ejemplo de funciones técnicas utilizando
# la librería TA para análisis técnico. Modificar al gusto de cada usuario.
# Además, crearemos las listas para añadir nuestros ticker que cumplan con el screener.
# 6. Following we will write some functions in order to retreive the technical 
# indicator from TA, add it to our dataframe and define logics for the signals.
# On top, we will create the list to fulfill with the tickers filtered.

b_out = []
cons = []
mcd_up = []
mcd_up0 = []
mcd_d = []
mcd_d0 = []
bb_up  = []
already_bb_up = []
bb_d  = []
already_bb_d = []
rsi_d = []
on_rsi_d = []
on_rsi_up = []
rsi_up = []
rsi_bf_d = []
rsi_bf_up = []

def MACD_signal_up(df):
	"""
	This Function will analyze the SIGNAL UP on the MACD of the asset
	Asset MACD on Crossover SIGNAL and Asset MACD Crossover below 0

	"""	
	from ta.trend import MACD
	indicator_macd = MACD(df['Adj Close'])
	df['MACD'] = indicator_macd.macd()
	df['Signal']= indicator_macd.macd_signal()
	df['MACD Histogram']= indicator_macd.macd_diff()
	df['Below_0_Crossover_MACD_Signal'] = False
	df['Simple_Crossover_MACD_Signal'] = False

	# MACD Crossover logics
	if (df[-2:]['MACD'].values[0] <= df[-1:]['MACD'].values[0]) and (df[-2:]['MACD'].values[0] <= df[-2:]['Signal'].values[0]) and (df[-1:]['MACD'].values[0]>=df[-1:]['Signal'].values[0]):
		

		# MACD crossover AND Below 0	
		if (df[-2:]['MACD'].values[0] <= df[-1:]['MACD'].values[0]) and (df[-2:]['MACD'].values[0] <= df[-2:]['Signal'].values[0]) and (df[-1:]['MACD'].values[0]>=df[-1:]['Signal'].values[0]) and df[-1:]['MACD'].values[0]<= 0:
			mcd_up0.append(symbol)
			df['Below_0_Crossover_MACD_Signal'][-1] = True
		else:
			mcd_up.append(symbol)
			df['Simple_Crossover_MACD_Signal'][-1] = True
			df['Below_0_Crossover_MACD_Signal'][-1] = False
			return True
	return False

def MACD_signal_down(df):
	"""
	This Function will analyze the SIGNAL DOWN on the MACD asset
	Asset MACD on Crossunder SIGNAL and Asset MACD Crossunder above 0

	"""	
	from ta.trend import MACD
	indicator_macd = MACD(df['Adj Close'])
	df['MACD'] = indicator_macd.macd()
	df['Signal']= indicator_macd.macd_signal()
	df['MACD Histogram']= indicator_macd.macd_diff()
	df['Simple_Crossdown_MACD_Signal'] = False
	df['Above_0_Crossunder_MACD_Signal'] = False

	#	MACD croosunder
	if (df[-2:]['MACD'].values[0] >= df[-1:]['MACD'].values[0]) and (df[-2:]['MACD'].values[0] >= df[-2:]['Signal'].values[0]) and (df[-1:]['MACD'].values[0]<=df[-1:]['Signal'].values[0]):
		# MACD crossunder AND above 0	
		if (df[-2:]['MACD'].values[0] >= df[-1:]['MACD'].values[0]) and (df[-2:]['MACD'].values[0] >= df[-2:]['Signal'].values[0]) and (df[-1:]['MACD'].values[0]<=df[-1:]['Signal'].values[0]) and df[-1:]['MACD'].values[0]>= 0:
			mcd_d0.append(symbol)
			df['Above_0_Crossunder_MACD_Signal'][-1] = True
			
		else:
			mcd_d.append(symbol)
			df['Simple_Crossdown_MACD_Signal'][-1] = True
			df['Above_0_Crossunder_MACD_Signal'][-1] = False
			return True
		return False

def Bollinger_signal_up(df, window=20, window_dev=2):
	"""
	This Function will analyze the Bollinger UP on the asset
	Bollinger Up signal, Asset already above upper Bollinger Band

	"""	

	from ta.volatility import BollingerBands
	indicator_bb = BollingerBands(df["Adj Close"], 20, 2)
	df['bb_bbm'] = indicator_bb.bollinger_mavg()
	df['bb_bbh'] = indicator_bb.bollinger_hband()
	df['bb_bbl'] = indicator_bb.bollinger_lband()
	df['bb_bbhi'] = indicator_bb.bollinger_hband_indicator()
	df['bb_bbli'] = indicator_bb.bollinger_lband_indicator()
	df['Boll_UP'] = False
	df['Boll_UP2']= False

	# Asset on Upper Bollinger Band Signal
	if (df[-2:]['bb_bbhi'].values[0] == 0) and (df[-1:]['bb_bbhi'].values[0] == 1):
		df['Boll_UP'][-1] = True
		bb_up.append(symbol)
		return True
	
	# Asset already avobe Upper Bollinger Band
	elif (df[-2:]['bb_bbhi'].values[0] == 1) and (df[-1:]['bb_bbhi'].values[0] == 1):
		df['Boll_UP2'][-1] = True
		already_bb_up.append(symbol)
		return True
	return False

def Bollinger_signal_down(df, window=20, window_dev=2):
	"""
	This Function will analyze the Bollinger DOWN on the asset
	Bollinger down signal, Asset already below lower Bollinger Band

	"""	

	from ta.volatility import BollingerBands
	indicator_bb = BollingerBands(df["Adj Close"], 20, 2)
	df['bb_bbm'] = indicator_bb.bollinger_mavg()
	df['bb_bbh'] = indicator_bb.bollinger_hband()
	df['bb_bbl'] = indicator_bb.bollinger_lband()
	df['bb_bbhi'] = indicator_bb.bollinger_hband_indicator()
	df['bb_bbli'] = indicator_bb.bollinger_lband_indicator()
	df['Boll_Down'] = False
	df['Boll_Down2']= False

	# Asset on Signal Lower Bollinger Band
	if (df[-2:]['bb_bbli'].values[0] == 0) and (df[-1:]['bb_bbli'].values[0] == 1):	
		bb_d.append(symbol)
		df['Boll_Down'][-1]= True
		return True
	
	# Asset already below lower Bollinger
	elif (df[-2:]['bb_bbli'].values[0] == 1) and (df[-1:]['bb_bbli'].values[0] == 1):
		already_bb_d.append(symbol)	
		df['Boll_Down2'][-1]= True
		return True
	return False

def RSI_signal_up(df, window = 14):
	"""
	This Function will analyze the SIGNAL UP on the RSI asset
	Overbought signal, Asset already Overbought and asset back to range from Overbought

	"""	

	from ta.momentum import RSIIndicator
	indicator_rsi= RSIIndicator(df['Adj Close'], window= 14)
	df['RSI'] = indicator_rsi.rsi()
	df['RSI_Overbought'] = False
	
	# Asset back to range 70-30 from Overbought
	if (df[-2:]['RSI'].values[0] >= 70) and (df[-1:]['RSI'].values[0] <= 70):
		rsi_bf_up.append(symbol)

	# Asset on RSI > 70		
	if (df[-1:]['RSI'].values[0] >= 70):
		on_rsi_up.append(symbol)
		df['RSI_Overbought'][-1] = True
		
		# RSI Overbought SIGNAL	
		if (df[-2:]['RSI'].values[0] <= 70) and (df[-1:]['RSI'].values[0] >= 70):
			rsi_up.append(symbol)
			return True
	return False

def RSI_signal_down(df, window= 14):
	"""
	This Function will analyze the SIGNAL DOWN on the RSI asset
	Oversold signal, Asset already oversold and asset back to range from oversold

	"""	

	from ta.momentum import RSIIndicator
	indicator_rsi= RSIIndicator(df['Adj Close'], window= 14)
	df['RSI'] = indicator_rsi.rsi()
	df['RSI_Oversold'] = False
	
	# Asset back to range 30-70 from Oversold
	if (df[-2:]['RSI'].values[0] <= 30) and (df[-1:]['RSI'].values[0] >= 30):
		rsi_bf_d.append(symbol)

	# Asset on RSI < 30
	if (df[-1:]['RSI'].values[0] <= 30):
		on_rsi_d.append(symbol)
		df['RSI_Oversold'][-1] = True
		
		# RSI just crossed down SIGNAL
		if (df[-2:]['RSI'].values[0] >= 30) and (df[-1:]['RSI'].values[0] <= 30):
			rsi_d.append(symbol)
			return True
	return False


# Price action Functions (candlesticks patterns, consolidations, breakouts etc.)

def consolidating_signal(df, perc = 3.5):
	"""
	This Function will analyze the asset is consolidating within the perc range.
	Ex: perc =3.5 means the closing price within the last 15 sessions, hasn't changed 
	further than 3.5%

	"""	
	range_of_candlesticks= df[-15:]
	max_close_price = range_of_candlesticks['Adj Close'].max()
	min_close_price = range_of_candlesticks['Adj Close'].min()
	threshold_detection = 1 - (perc / 100)
	if min_close_price > (max_close_price * threshold_detection):
		cons.append(symbol)
		return True
	return False

def breaking_out_signal(df, perc=1,):
	"""
	This Function will analyze the an asset which is coming out from a consolidation
	period.
	
	[perc] = will be the threshold in % for the closing price to determinate if the asset is 
	under consolidation.

	On the example perc = 1, the asset will be closing within 1% range on the last 15 sessions and then
	on current candle is breaking out.

	"""	
	last_close = df[-1:]['Adj Close'].values[0]
	if consolidating_signal(df[:-1], perc = perc):
		recent_close = df[-16:-1]
		if last_close > recent_close['Adj Close'].max():
			b_out.append(symbol)
			return True
	return False


# START SCREENER  with our data stored in .CSV format
print(f'--------- GENERAL MARKET SCREENER in {country} for {len(stocks)} assets: data analyzed from {start} until {today} --------\n')

for filename in os.listdir(path):
	df = pd.read_csv(path+f'\{filename}')
	symbol = filename.split(".")[0]
	if breaking_out_signal(df, 3):
		pass
	if consolidating_signal(df, perc=2):
		pass
	if RSI_signal_up(df):
		pass
	if RSI_signal_down(df):
		pass
	if MACD_signal_up(df):
		pass
	if MACD_signal_down(df):
		pass
	if Bollinger_signal_up(df):
		pass
	if Bollinger_signal_down(df):
		pass


# OUTPUT

print('--- BOLLINGER ANALYSIS --- \n')
print(f'The stocks on SIGNAL BOLLINGER UP are:\n==> {bb_up}\n')
print(f'The stocks are already in BOLLINGER UP:\n==> {already_bb_up}\n')
print(f'The stocks on SIGNAL BOLLINGER DOWN are:\n==> {bb_d}\n')
print(f'The stocks are already in BOLLINGER_DOWN:\n==> {already_bb_d}\n')

print('--- MACD ANALYSIS --- \n')
print(f'The stocks on MACD SIGNAL UP are:\n==> {mcd_up}\n')
print(f'The stocks on MACD SIGNAL UP BELOW 0 are:\n==> {mcd_up0}\n')
print(f'The stocks on MACD SIGNAL DOWN are:\n==> {mcd_d} \n')
print(f'The stocks on MACD SIGNAL DOWN above 0 are:\n==> {mcd_d0}\n')

print('--- RSI ANALYSIS --- \n')
print(f'The stocks on OVERBOUGHT SIGNAL [RSI] are:\n==> {rsi_up}\n')
print(f'The stocks on OVERSOLD SIGNAL [RSI] are:\n==> {rsi_d}\n')
print(f'The stocks went to RANGE from OVERSOLD are:\n==> {rsi_bf_d}\n')
print(f'The stocks went to RANGE from OVERBOUGHT are:\n==> {rsi_bf_up}\n')
print(f'The stocks on OVERBOUGHT [RSI] are:\n==> {on_rsi_up}\n')
print(f'The stocks on OVERSOLD [RSI] are:\n==> {on_rsi_d}\n')

print('--- PRICE ACTION ANALYSIS --- \n')
print(f'The stocks on CONSOLIDATION are:\n==> {cons}\n')
print(f'The stocks on BREAKOUT are:\n==> {b_out}\n')

# Delete stored information UNCOMMENT IF WANT TO REMOVE AFTER SCREENER
shutil.rmtree(path)