statsmodels.tsa.statespace.kalman_filter.FilterResults

class statsmodels.tsa.statespace.kalman_filter.FilterResults(model)

Results from applying the Kalman filter to a state space model.

Parameters:

model (Representation) – A Statespace representation

nobs

Number of observations.

Type:

int

nobs_diffuse

Number of observations under the diffuse Kalman filter.

Type:

int

k_endog

The dimension of the observation series.

Type:

int

k_states

The dimension of the unobserved state process.

Type:

int

k_posdef

The dimension of a guaranteed positive definite covariance matrix describing the shocks in the measurement equation.

Type:

int

dtype

Datatype of representation matrices

Type:

dtype

prefix

BLAS prefix of representation matrices

Type:

str

shapes

A dictionary recording the shapes of each of the representation matrices as tuples.

Type:

dictionary of name,tuple

endog

The observation vector.

Type:

ndarray

design

The design matrix, \(Z\).

Type:

ndarray

obs_intercept

The intercept for the observation equation, \(d\).

Type:

ndarray

obs_cov

The covariance matrix for the observation equation \(H\).

Type:

ndarray

transition

The transition matrix, \(T\).

Type:

ndarray

state_intercept

The intercept for the transition equation, \(c\).

Type:

ndarray

selection

The selection matrix, \(R\).

Type:

ndarray

state_cov

The covariance matrix for the state equation \(Q\).

Type:

ndarray

missing

An array of the same size as endog, filled with boolean values that are True if the corresponding entry in endog is NaN and False otherwise.

Type:

array of bool

nmissing

An array of size nobs, where the ith entry is the number (between 0 and k_endog) of NaNs in the ith row of the endog array.

Type:

array of int

time_invariant

Whether or not the representation matrices are time-invariant

Type:

bool

initialization

Kalman filter initialization method.

Type:

str

initial_state

The state vector used to initialize the Kalamn filter.

Type:

array_like

initial_state_cov

The state covariance matrix used to initialize the Kalamn filter.

Type:

array_like

initial_diffuse_state_cov

Diffuse state covariance matrix used to initialize the Kalamn filter.

Type:

array_like

filter_method

Bitmask representing the Kalman filtering method

Type:

int

inversion_method

Bitmask representing the method used to invert the forecast error covariance matrix.

Type:

int

stability_method

Bitmask representing the methods used to promote numerical stability in the Kalman filter recursions.

Type:

int

conserve_memory

Bitmask representing the selected memory conservation method.

Type:

int

filter_timing

Whether or not to use the alternate timing convention.

Type:

int

tolerance

The tolerance at which the Kalman filter determines convergence to steady-state.

Type:

float

loglikelihood_burn

The number of initial periods during which the loglikelihood is not recorded.

Type:

int

converged

Whether or not the Kalman filter converged.

Type:

bool

period_converged

The time period in which the Kalman filter converged.

Type:

int

filtered_state

The filtered state vector at each time period.

Type:

ndarray

filtered_state_cov

The filtered state covariance matrix at each time period.

Type:

ndarray

predicted_state

The predicted state vector at each time period.

Type:

ndarray

predicted_state_cov

The predicted state covariance matrix at each time period.

Type:

ndarray

forecast_error_diffuse_cov

Diffuse forecast error covariance matrix at each time period.

Type:

ndarray

predicted_diffuse_state_cov

The predicted diffuse state covariance matrix at each time period.

Type:

ndarray

kalman_gain

The Kalman gain at each time period.

Type:

ndarray

forecasts

The one-step-ahead forecasts of observations at each time period.

Type:

ndarray

forecasts_error

The forecast errors at each time period.

Type:

ndarray

forecasts_error_cov

The forecast error covariance matrices at each time period.

Type:

ndarray

llf_obs

The loglikelihood values at each time period.

Type:

ndarray

Methods

predict([start, end, dynamic])	In-sample and out-of-sample prediction for state space models generally
update_filter(kalman_filter)	Update the filter results
update_representation(model[, only_options])	Update the results to match a given model

Properties

kalman_gain	Kalman gain matrices
standardized_forecasts_error	Standardized forecast errors