import pandas as pd import numpy as np import seaborn as sns #sns.get_dataset_names() from pandas_datareader import data from sklearn.model_selection import train_test_split import time import datetime

ticker = 'FB2A.BE' period1 = int(time.mktime(datetime.datetime(2010, 1, 1, 23, 59).timetuple())) period2 = int(time.mktime(datetime.datetime(2022, 2, 1, 23, 59).timetuple())) interval = '1d' query_string = f'https://query1.finance.yahoo.com/v7/finance/download/{ticker}?period1={period1}&period2={period2}&interval={interval}&events=history&includeAdjustedClose=true' data = pd.read_csv(query_string) data

data.info()

data['highest hight'] = data['High'].rolling(window=10).max() data['lowest low'] = data['Low'].rolling(window=10).min() data.info()

data

data['trigger'] = np.where(data['High']==data['highest hight'],1,np.nan) data['trigger'] = np.where(data['Low']==data['lowest low'],0,data['trigger'])

data['position'] = data['trigger'].ffill().fillna(-1) data

df = data.drop(['Date','Adj Close','Volume','highest hight', 'lowest low', 'trigger'],axis=1)

df

X = df.drop('position',axis=1)

X

y = df['position']

y

from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=0)

from sklearn.preprocessing import StandardScaler sc = StandardScaler() X_train = sc.fit_transform(X_train) X_test = sc.transform(X_test)

from sklearn import svm svm_reg = svm.SVR() regressor = svm_reg.fit(X_train, y_train)

regressor

y_pre = regressor.predict(X_test)

import pickle

pkl_filename = "Facebook.pkl" with open(pkl_filename, 'wb') as file: pickle.dump(regressor, file)

with open(pkl_filename, 'rb') as file: model = pickle.load(file)

x_in = np.asarray([320.47,320.88,304.00,304.79]).reshape(1,-1) predicts = regressor.predict(x_in) predicts[0]