Hands-On Machine Learning for Algorithmic Trading

Preparing the data

We apply the same transformation—annual difference for both series, prior log-transform for industrial production—to achieve stationarity (see Chapter 8, Time Series Models, for details), as shown here:

df_transformed = pd.DataFrame({'ip': np.log(df.ip).diff(12),                    'sentiment': df.sentiment.diff(12)}).dropna()

The create_multivariate_rnn_data function transforms a dataset of several time series into the shape required by the Keras RNN layers, namely n_samples x window_size x n_series, as follows:

def create_multivariate_rnn_data(data, window_size):    y = data[window_size:]    n = data.shape[0]    X = np.stack([data[i: j] for i, j in enumerate(range(window_size, n))], axis=0)    return X, y

We will use window_size of 24 months and obtain ...