Hi,
I'm trying to smooth the data from the history function, using the kalman filter. But the kalmanfilter function only takes an array (I think), but the data from the history function is a dataframe.
I thought I would loop over all the sids and call the filter, but I'm not sure how to implement it. Anyone know how to implement this=?
prices = np.log(history(context.lookback, '1d', 'price').dropna(axis=1))
# Construct a Kalman filter
kf = KalmanFilter(transition_matrices = [1],
observation_matrices = [1],
initial_state_mean = 0,
initial_state_covariance = 1,
observation_covariance=1,
transition_covariance=.01)
for column in prices.T:
mean_prices, _ = kf.filter(column)