moderndid.NPIVResult#

class moderndid.NPIVResult(h: ndarray, h_lower: ndarray | None, h_upper: ndarray | None, deriv: ndarray, h_lower_deriv: ndarray | None, h_upper_deriv: ndarray | None, beta: ndarray, asy_se: ndarray, deriv_asy_se: ndarray, cv: float | None, cv_deriv: float | None, residuals: ndarray, j_x_degree: int, j_x_segments: int, k_w_degree: int, k_w_segments: int, args: dict)[source]#

Container for nonparametric instrumental variables estimation results.

Attributes:

hnumpy.ndarray: Estimated structural function \(\hat{h}_J(x)\) at evaluation points.
h_lowernumpy.ndarray or None: Lower uniform confidence band for \(h_0\).
h_uppernumpy.ndarray or None: Upper uniform confidence band for \(h_0\).
derivnumpy.ndarray: Estimated derivative \(\partial^a \hat{h}_J(x)\) at evaluation points.
h_lower_derivnumpy.ndarray or None: Lower uniform confidence band for \(\partial^a h_0\).
h_upper_derivnumpy.ndarray or None: Upper uniform confidence band for \(\partial^a h_0\).
betanumpy.ndarray: Sieve coefficient vector \(\hat{c}_J\).
asy_senumpy.ndarray: Pointwise asymptotic standard errors \(\hat{\sigma}_J(x)\).
deriv_asy_senumpy.ndarray: Pointwise asymptotic standard errors \(\hat{\sigma}_J^a(x)\) for derivatives.
cvfloat or None: Bootstrap critical value \(z_{1-\alpha}^*\) for function UCBs.
cv_derivfloat or None: Bootstrap critical value \(z_{1-\alpha}^{a*}\) for derivative UCBs.
residualsnumpy.ndarray: TSLS residuals \(\hat{u}_{i,J} = Y_i - \hat{h}_J(X_i)\).
j_x_degreeint: Degree of B-spline basis for \(X\).
j_x_segmentsint: Number of segments for \(X\) basis.
k_w_degreeint: Degree of B-spline basis for \(W\).
k_w_segmentsint: Number of segments for \(W\) basis.
argsdict: Diagnostic information. When data-driven selection is used, includes j_x_seg, k_w_seg, j_hat_max, theta_star, and other selection diagnostics from the Lepski procedure.

Methods

`count`(value, /)	Return number of occurrences of value.
`index`(value[, start, stop])	Return first index of value.

Attributes

`args`	Diagnostic information and selection diagnostics.
`asy_se`	Pointwise asymptotic standard errors for the structural function.
`beta`	Sieve coefficient vector.
`cv`	Bootstrap critical value for function uniform confidence bands.
`cv_deriv`	Bootstrap critical value for derivative uniform confidence bands.
`deriv`	Estimated derivative at evaluation points.
`deriv_asy_se`	Pointwise asymptotic standard errors for derivatives.
`h`	Estimated structural function at evaluation points.
`h_lower`	Lower uniform confidence band for the structural function.
`h_lower_deriv`	Lower uniform confidence band for the derivative.
`h_upper`	Upper uniform confidence band for the structural function.
`h_upper_deriv`	Upper uniform confidence band for the derivative.
`j_x_degree`	Degree of B-spline basis for X.
`j_x_segments`	Number of segments for X basis.
`k_w_degree`	Degree of B-spline basis for W.
`k_w_segments`	Number of segments for W basis.
`residuals`	TSLS residuals.