Есть библиотека для расчета индикаторов по техническому анализу. Ниже пример (не все идеально, как ориентир). Ссылка на библиотеку https://github.com/mrjbq7/ta-lib

#=====================================================
# Импорт библиотек
#=====================================================
from iexfinance.stocks import get_historical_data
import datetime
import pandas as pd
from pandas import DataFrame
import talib as ta
import matplotlib.pyplot as plt
import seaborn as sns
import psycopg2
import csv
from sqlalchemy import create_engine
import json
import requests
from sys import argv
from scipy.signal import argrelextrema
import numpy as np

#=====================================================
# ЗАГРУЗКА ДАННЫХ ИЗ API
#=====================================================
symbol = argv[1]
interval = argv[2]
urltext = 'https://api.binance.com/api/v1/klines?symbol='+symbol+'&interval='+interval
result_req = requests.get(urltext)

result_json = result_req.json()

# Получаем наименования колонок
array_data = []
for row in result_json:
    array_data.append({
                        'open_time':row[0],
                        'date_time':datetime.datetime.fromtimestamp(row[0] / 1e3).strftime('%Y-%m-%d %H:%M:%S'), #utcfromtimestamp fromtimestamp #utcfromtimestamp(ts).strftime('%Y-%m-%d %H:%M:%S')
            'dateformat':datetime.datetime.fromtimestamp(row[0] / 1e3).strftime('%Y.%m.%d %H:%M'),
                        'open':(row[1]),
                        'high':(row[2]),
                        'low':(row[3]),
                        'close':(row[4]),
                        'volume':(row[5])
                        })
    

# Формируем массив data
data = pd.DataFrame(array_data)

multipleCoeff = 1000
data['open']=pd.to_numeric(data['open']).mul(multipleCoeff)
data['high']=pd.to_numeric(data['high']).mul(multipleCoeff)
data['low']=pd.to_numeric(data['low']).mul(multipleCoeff)
data['close']=pd.to_numeric(data['close']).mul(multipleCoeff)
data['volume']=pd.to_numeric(data['volume']).mul(multipleCoeff)

data.index = pd.to_datetime(data.date_time)

#==============================================
# ПОЛУЧАЕМ ПАРАМЕТРЫ ИЗ БАЗЫ ДАННЫХ
#==============================================

# 1. Подключаемся к базе данных PGSQL
conn = psycopg2.connect(dbname='dbname', user='username', 
                        password='your_password', host='96.154.12.134')
# 2. Получаем данные, кладем их в курсор
cursor = conn.cursor()
cursor.execute('select trim(parameter) parameter,value from ddt.calc_settings')

# 3. Пишем paramData[]
paramData = []
for paramRow in cursor:
    paramData.append(paramRow)

# 4. Закрываем курсор
cursor.close()
conn.close()

# 5. Устанавливаем переменные
for row in enumerate(paramData):
    exec("%s=%s" % (row[1][0],row[1][1]))


#=====================================================
# Расчет
#=====================================================

#Фрактал
n_min_max = 5
data['Fractal_UP'] = data.iloc[argrelextrema(data.low.values, np.less_equal, order=n_min_max)[0]]['low']
data['Fractal_Down'] = data.iloc[argrelextrema(data.high.values, np.greater_equal, order=n_min_max)[0]]['high']

# Обнуляем последний элемент массива по фракталу
data['Fractal_UP'].iloc[-1] = np.nan
data['Fractal_Down'].iloc[-1] = np.nan

# Simple Moving Average
data['SMA'] = ta.SMA(data.close, timeperiod = SMA_TimePeriod)

# Exponential Moving Average
data['EMA'] = ta.EMA(data.close, timeperiod = EMA_TimePeriod)

# Bollinger Bands
data['upper_band'], data['middle_band'], data['lower_band'] = ta.BBANDS(data.close, timeperiod = Bollinger_Bands_TimePeriod)

# MACD
data['macd'], data['macdsignal'], data['macdhist'] = ta.MACD(data.close, fastperiod=MACD_fastperiod, slowperiod=MACD_slowperiod, signalperiod=MACD_signalperiod)

# RSI
data['RSI'] = ta.RSI(data.close, timeperiod=RSI_TimePeriod)

# OBV (On Balance Volume)
data['OBV'] = ta.OBV(data.close, data.volume)

#Alligator SMA
data['SMA_green'] = ta.SMA(data.close, timeperiod = SMA_Green_TimePeriod)
data['SMA_red'] = ta.SMA(data.close, timeperiod = SMA_Red_TimePeriod)
data['SMA_blue'] = ta.SMA(data.close, timeperiod = SMA_Blue_TimePeriod)

#Alligator EMA
data['EMA_green'] = ta.EMA(data.close, timeperiod = EMA_Green_TimePeriod)
data['EMA_red'] = ta.EMA(data.close, timeperiod = EMA_Red_TimePeriod)
data['EMA_blue'] = ta.EMA(data.close, timeperiod = EMA_Blue_TimePeriod)

#ADX
data['ADX'] = ta.ADX(data.high, data.low, data.close, timeperiod=ADX_TimePeriod)
data['MINUS_DI'] = ta.MINUS_DI(data.high, data.low, data.close, timeperiod=ADX_TimePeriod)
data['PLUS_DI'] = ta.PLUS_DI(data.high, data.low, data.close, timeperiod=ADX_TimePeriod)

data['TEMA'] = ta.TEMA(data.close, timeperiod=TEMA_TimePeriod)

# Убираем масштабирование
data['open'] = pd.to_numeric(data['open'])/multipleCoeff
data['high'] = pd.to_numeric(data['high'])/multipleCoeff
data['low'] = pd.to_numeric(data['low'])/multipleCoeff
data['close'] = pd.to_numeric(data['close'])/multipleCoeff
data['upper_band'] = pd.to_numeric(data['upper_band'])/multipleCoeff
data['middle_band'] = pd.to_numeric(data['middle_band'])/multipleCoeff
data['lower_band'] = pd.to_numeric(data['lower_band'])/multipleCoeff
data['volume'] = pd.to_numeric(data['volume'])/multipleCoeff
data['OBV'] = pd.to_numeric(data['OBV'])/multipleCoeff
data['SMA'] = pd.to_numeric(data['SMA'])/multipleCoeff
data['EMA'] = pd.to_numeric(data['EMA'])/multipleCoeff
data['macd'] = pd.to_numeric(data['macd'])/multipleCoeff
data['macdsignal'] = pd.to_numeric(data['macdsignal'])/multipleCoeff
data['macdhist'] = pd.to_numeric(data['macdhist'])/multipleCoeff

data['SMA_green'] = pd.to_numeric(data['SMA_green'].shift(periods=SMA_Green_Offset))/multipleCoeff
data['SMA_red'] = pd.to_numeric(data['SMA_red'].shift(periods=SMA_Red_Offset))/multipleCoeff
data['SMA_blue'] = pd.to_numeric(data['SMA_blue'].shift(periods=SMA_Blue_Offset))/multipleCoeff

data['EMA_green'] = pd.to_numeric(data['EMA_green'].shift(periods=EMA_Green_Offset))/multipleCoeff
data['EMA_red'] = pd.to_numeric(data['EMA_red'].shift(periods=EMA_Red_Offset))/multipleCoeff
data['EMA_blue'] = pd.to_numeric(data['EMA_blue'].shift(periods=EMA_Blue_Offset))/multipleCoeff

data['TEMA'] = pd.to_numeric(data['TEMA'])/multipleCoeff

data['Fractal_UP'] = pd.to_numeric(data['Fractal_UP'])/multipleCoeff
data['Fractal_Down'] = pd.to_numeric(data['Fractal_Down'])/multipleCoeff

#=====================================================
# JSON
#=====================================================

jsonstring = data.to_json(orient='records')
print(jsonstring)

Вызвать скрипт можно из PHP, вот скрипт:

<?php	

  //Получаем переменные из URL запроса
  if (isset($_GET['symbol']) and isset($_GET['interval'])) {
    $symbol   = htmlspecialchars($_GET["symbol"]);
    $interval   = htmlspecialchars($_GET["interval"]);
  }
  else {
    err();
  }        

  echo exec("python3 /home/admin/web/domain-name.ru/public_html/panel/98hlkjg987gtkbjbIUYG98iub/python_api/5_indicators.py ".$symbol." ".$interval);
                                                                                                                                                     
?>

Сообщение Пример скрипта для расчета индикаторов TA_LIB Python появились сначала на Python 3 | Data Science | Нейронные сети | AI - Искусственный Интеллект.