# statsmodels.tsa.statespace.kalman_filter.FilterResults¶

class statsmodels.tsa.statespace.kalman_filter.FilterResults(model)[source]

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

Parameters:
modelRepresentation

A Statespace representation

Attributes:
nobsint

Number of observations.

nobs_diffuseint

Number of observations under the diffuse Kalman filter.

k_endogint

The dimension of the observation series.

k_statesint

The dimension of the unobserved state process.

k_posdefint

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

dtypedtype

Datatype of representation matrices

prefixstr

BLAS prefix of representation matrices

shapesdictionary of name,tuple

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

endogndarray

The observation vector.

designndarray

The design matrix, $$Z$$.

obs_interceptndarray

The intercept for the observation equation, $$d$$.

obs_covndarray

The covariance matrix for the observation equation $$H$$.

transitionndarray

The transition matrix, $$T$$.

state_interceptndarray

The intercept for the transition equation, $$c$$.

selectionndarray

The selection matrix, $$R$$.

state_covndarray

The covariance matrix for the state equation $$Q$$.

missingarray of bool

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.

nmissingarray of int

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.

time_invariantbool

Whether or not the representation matrices are time-invariant

initializationstr

Kalman filter initialization method.

initial_statearray_like

The state vector used to initialize the Kalamn filter.

initial_state_covarray_like

The state covariance matrix used to initialize the Kalamn filter.

initial_diffuse_state_covarray_like

Diffuse state covariance matrix used to initialize the Kalamn filter.

filter_methodint

Bitmask representing the Kalman filtering method

inversion_methodint

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

stability_methodint

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

conserve_memoryint

Bitmask representing the selected memory conservation method.

filter_timingint

Whether or not to use the alternate timing convention.

tolerancefloat

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

loglikelihood_burnint

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

convergedbool

Whether or not the Kalman filter converged.

period_convergedint

The time period in which the Kalman filter converged.

filtered_statendarray

The filtered state vector at each time period.

filtered_state_covndarray

The filtered state covariance matrix at each time period.

predicted_statendarray

The predicted state vector at each time period.

predicted_state_covndarray

The predicted state covariance matrix at each time period.

forecast_error_diffuse_covndarray

Diffuse forecast error covariance matrix at each time period.

predicted_diffuse_state_covndarray

The predicted diffuse state covariance matrix at each time period.

kalman_gainndarray

Kalman gain matrices

forecastsndarray

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

forecasts_errorndarray

The forecast errors at each time period.

forecasts_error_covndarray

The forecast error covariance matrices at each time period.

llf_obsndarray

The loglikelihood values at each time period.

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