Avertissement: Ce code vise à partager publiquement des exemples de code. Nous déclinons toutes responsabilité en cas de perte fiancière.
Créer un dossier nommé "csv".
Dans le dossier "csv", mettre ces fichiers. Ceux-ci sont les données de prix de la bourse japonaise.
Télécharger les données: https://github.com/shunakanishi/japanese_stockprice
Maintenant, aller dans le premier répertoire et créer un fichier de python 3.
Préparer le modèle et les paramètres dans le dossier "model" -> "stockprice".
Dans le "model":
Dans le "stockprice":
Si vous n'avons pas le modèle et les paramètres, aller voir ce post et créer le modèle et les paramètres d'abord.
Insérer le code suivant dans le fichier "stockprice.py":
#-*- coding: utf-8 -*-
import numpy
import pandas
import matplotlib.pyplot as plt
from sklearn import preprocessing
from keras.models import Sequential
from keras.models import model_from_json
from keras.layers.core import Dense, Activation
from keras.layers.recurrent import LSTM
import keras.backend.tensorflow_backend as KTF
import os.path
class Prediction :
def __init__(self):
self.length_of_sequences = 10
self.in_out_neurons = 1
self.hidden_neurons = 300
def load_data(self, data, n_prev=10):
X, Y = [], []
for i in range(len(data) - n_prev):
X.append(data.iloc[i:(i+n_prev)].as_matrix())
Y.append(data.iloc[i+n_prev].as_matrix())
retX = numpy.array(X)
retY = numpy.array(Y)
return retX, retY
def create_model(self, f_model, model_filename, weights_filename) :
print(os.path.join(f_model,model_filename))
if os.path.isfile(os.path.join(f_model,model_filename)):
print('Saved parameters found. I will use this file...')
json_string = open(os.path.join(f_model, model_filename)).read()
model = model_from_json(json_string)
model.summary()
model.compile(loss="mape", optimizer="adam")
model.load_weights(os.path.join(f_model,weights_filename))
else:
print('Saved parameters Not found. Please prepare model and parameters.')
model = None
return model
if __name__ == "__main__":
f_log = './log'
f_model = './model/stockprice'
model_filename = 'stockprice_model.json'
yaml_filename = 'stockprice_model.yaml'
weights_filename = 'stockprice_model_weights.hdf5'
prediction = Prediction()
# Data
data = None
for year in range(2007, 2017):
data_ = pandas.read_csv('csv/indices_I101_1d_' + str(year) + '.csv')
data = data_ if (data is None) else pandas.concat([data, data_])
data.columns = ['date', 'open', 'high', 'low', 'close']
data['date'] = pandas.to_datetime(data['date'], format='%Y-%m-%d')
# Data of closing price
data['close'] = preprocessing.scale(data['close'])
data = data.sort_values(by='date')
data = data.reset_index(drop=True)
data = data.loc[:, ['date', 'close']]
# 100% of the data is used as test data.
# split_pos = int(len(data) * 0.8)
# x_train, y_train = prediction.load_data(data[['close']].iloc[0:split_pos], prediction.length_of_sequences)
x_test, y_test = prediction.load_data(data[['close']], prediction.length_of_sequences)
old_session = KTF.get_session()
model = prediction.create_model(f_model, model_filename, weights_filename)
predicted = model.predict(x_test)
result = pandas.DataFrame(predicted)
result.columns = ['predict']
result['actual'] = y_test
result.plot()
plt.show()
import numpy
import pandas
import matplotlib.pyplot as plt
from sklearn import preprocessing
from keras.models import Sequential
from keras.models import model_from_json
from keras.layers.core import Dense, Activation
from keras.layers.recurrent import LSTM
import keras.backend.tensorflow_backend as KTF
import os.path
class Prediction :
def __init__(self):
self.length_of_sequences = 10
self.in_out_neurons = 1
self.hidden_neurons = 300
def load_data(self, data, n_prev=10):
X, Y = [], []
for i in range(len(data) - n_prev):
X.append(data.iloc[i:(i+n_prev)].as_matrix())
Y.append(data.iloc[i+n_prev].as_matrix())
retX = numpy.array(X)
retY = numpy.array(Y)
return retX, retY
def create_model(self, f_model, model_filename, weights_filename) :
print(os.path.join(f_model,model_filename))
if os.path.isfile(os.path.join(f_model,model_filename)):
print('Saved parameters found. I will use this file...')
json_string = open(os.path.join(f_model, model_filename)).read()
model = model_from_json(json_string)
model.summary()
model.compile(loss="mape", optimizer="adam")
model.load_weights(os.path.join(f_model,weights_filename))
else:
print('Saved parameters Not found. Please prepare model and parameters.')
model = None
return model
if __name__ == "__main__":
f_log = './log'
f_model = './model/stockprice'
model_filename = 'stockprice_model.json'
yaml_filename = 'stockprice_model.yaml'
weights_filename = 'stockprice_model_weights.hdf5'
prediction = Prediction()
# Data
data = None
for year in range(2007, 2017):
data_ = pandas.read_csv('csv/indices_I101_1d_' + str(year) + '.csv')
data = data_ if (data is None) else pandas.concat([data, data_])
data.columns = ['date', 'open', 'high', 'low', 'close']
data['date'] = pandas.to_datetime(data['date'], format='%Y-%m-%d')
# Data of closing price
data['close'] = preprocessing.scale(data['close'])
data = data.sort_values(by='date')
data = data.reset_index(drop=True)
data = data.loc[:, ['date', 'close']]
# 100% of the data is used as test data.
# split_pos = int(len(data) * 0.8)
# x_train, y_train = prediction.load_data(data[['close']].iloc[0:split_pos], prediction.length_of_sequences)
x_test, y_test = prediction.load_data(data[['close']], prediction.length_of_sequences)
old_session = KTF.get_session()
model = prediction.create_model(f_model, model_filename, weights_filename)
predicted = model.predict(x_test)
result = pandas.DataFrame(predicted)
result.columns = ['predict']
result['actual'] = y_test
result.plot()
plt.show()
Et faire ce commande:
$ sudo python3 stockprice.py
