Linear utility functions

---

source

get_linear_model

 get_linear_model (X:numpy.ndarray, X_prime:numpy.ndarray,
                   sorted:bool=True)

Returns the linear model A such that X’ = AX

	Type	Default	Details
X	ndarray		data matrix (state_vars, time_steps)
X_prime	ndarray		shifted data matrix (state_vars, time_steps)
sorted	bool	True	sort eigenvalues and eigenvectors
Returns	Tuple		(eigenvalues, eigenvectors)

---

source

sort_eigvals_eigvecs

 sort_eigvals_eigvecs (eigvals:numpy.ndarray, eigvecs:numpy.ndarray)

	Type	Details
eigvals	ndarray	eigenvalues
eigvecs	ndarray	eigenvectors
Returns	Tuple	(eigenvalues, eigenvectors)

DMD Utils

---

source

fast_predict

 fast_predict (y:jax.Array, inv_modes:jax.Array, fwd_modes:jax.Array,
               eigs:jax.Array, lenght:int=None)

	Type	Default	Details
y	Array		gridsize
inv_modes	Array
fwd_modes	Array
eigs	Array
lenght	int	None

---

source

fit_dmd_to_sample

 fit_dmd_to_sample (x:jax.Array, r:int=50)

	Type	Default	Details
x	Array
r	int	50
Returns	DMD		(timesteps, gridpoints)

---

source

hankelize

 hankelize (u:jax.Array, d:int=2)

Utilities to replace weights

---

source

set_params_from_linear

 set_params_from_linear (params:Dict, single_eigvals:jax.Array,
                         single_lstvecs:jax.Array, model:str)

	Type	Details
params	Dict
single_eigvals	Array
single_lstvecs	Array
model	str	“lru” or “koopman”
Returns	Dict

---

source

get_linear_approximation

 get_linear_approximation (y:jax.Array, r:int=50, method:str='dmd')

	Type	Default	Details
y	Array		(time_steps, grid_size)
r	int	50
method	str	dmd	“dmd”, “full”, “analytical”
Returns	Tuple