scspill_sim_dgp

simulate.scspill_sim_dgp(
    T0,
    T1,
    N,
    W,
    w,
    rho,
    sigma2,
    alpha,
    K=0,
    beta=None,
    seed=None,
    mu_tau=1.0,
    sd_tau=1.0,
    keep_internals=False,
)

Simulate one spillover panel from the paper’s SAR DGP.

Parameters

Name	Type	Description	Default
T0	int	Pre- and post-treatment lengths.	required
T1	int	Pre- and post-treatment lengths.	required
N	int	Number of control units.	required
W	np.ndarray	Raw control-to-control weights, `(N, N)` (row-normalized inside).	required
w	np.ndarray	Raw treated-to-control exposure, `(N,)` (sum-normalized inside).	required
rho	float	True spillover intensity.	required
sigma2	float	Error variance.	required
alpha	np.ndarray	True synthetic weights, `(N,)`.	required
K	int	Number of iid `N(0, 1)` covariates.	`0`
beta	np.ndarray	Covariate coefficients (required when `K > 0`).	`None`
seed	int or numpy.random.SeedSequence	Seed for `numpy.random.default_rng`.	`None`
mu_tau	float	Mean and sd of the per-period treatment effects.	`1.0`
sd_tau	float	Mean and sd of the per-period treatment effects.	`1.0`
keep_internals	bool	Keep the error draws on the returned object (for identity tests).	`False`

Returns

Name	Type	Description
	SimDGP

Raises

Name	Type	Description
	ScspillDataError	On shape mismatches or a near-singular SAR system.