Nonparametric Econometrics: The np Package

By T. Hayfield, Jeffrey S. Racine. 2008

912 citations recorded by the source corpus.

We describe the R np package via a series of applications that may be of interest to applied econometricians. The np package implements a variety of nonparametric and semiparametric kernel-based estimators that are popular among econometricians. There are also procedures for nonparametric tests of significance and consistent model specification tests for parametric mean regression models and parametric quantile regression models, among others. The np package focuses on kernel methods appropriate for the mix of continuous, discrete, and categorical data often found in applied settings. Data-driven methods of bandwidth selection are emphasized throughout, though we caution the user that data-driven bandwidth selection methods can be computationally demanding.

Subfields: Finance, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/a31342e1d9a674545c23e763f43c29381ea6556e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Finance

In Statistics and Probability

Citations in the corpus

On the spatial stickiness of UK new firm formation rates (2014, methodology): “Houston et al. (2008) critically examine whether Florida’s (2002) proposition that regions should attract members of the ‘creative class’(13) in order to enhance their economic success”
Conditional transformation models (2012, methodology): “Cross-validation was used to determine the bandwidths for the kernel-based methods, for details see Hayfield and Racine (2008). The boosting-based estimation of GAMLSS models (Mayr, Fenske, Hofner, Kneib, and Schmid 2012a) turned out to be more stable than the reference implementation (package gamls”
Deep IV: A Flexible Approach for Counterfactual Prediction (2017, methodology): “…addition to Deep IV, we considered a regular feed-forward network (FFNet) with the same architecture as our outcome network, a non-parametric IV polynomial kernel regression (NonPar, Darolles et al., 2011) using Hayfield et al. (2008)’s R implementation, and standard two-stage least squares (2SLS).”
Unobserved heterogeneity and endogeneity in nonparametric frontier estimation (2016, neutral): “9This can be obtained by the package ‘np’ in R, see Hayfield and Racine (2008).”
Converting High-Dimensional Regression to High-Dimensional Conditional Density Estimation (2017, methodology): “We use the R package np [23] with kd-trees and Epanechnikov kernels for computational efficiency [19, 25].”
POSTERIOR CONSISTENCY IN CONDITIONAL DENSITY ESTIMATION BY COVARIATE DEPENDENT MIXTURES (2011, methodology): “…to Appendix C. Additionally, we compare the KSB model with the nonparametric kernel smoothing method of Hall, Racine, and Li (2004) implemented by Hayfield and Racine (2008) in the publicly available R package np. Figure 2 shows that the estimation results for the DGP (6.3) from both approaches…”
Natural disasters and labor markets (2017, methodology): “The nonparametric density estimates are produced using the np package in R (Hayfield and Racine, 2008) and derived using a second order Epanechnikov as the kernel function; the smoothing parameter was selected via least squares crossvalidation as suggested by Rudemo (1982) and Bowman (1984) which mi”
What explains the performance of students in a heterogeneous environment? Conditional efficiency estimation with continuous and discrete environmental variables (2013, methodology): “The code utilizes some features of the np package by Hayfield and Racine (2008).”
Identification and estimation of semiparametric two‐step models (2016, methodology): “We make use of the np package by Hayfield and Racine (2008) in the statistical computing environment R.”
An efficient semiparametric maxima estimator of the extremal index (2015, methodology): “…standard errors based on a sandwich estimator (Appendix B) and bootstrap standard errors based on 100 stationary block bootstrap resamples (Politis and Romano, 1994) with optimal block length chosen using Patton et al (2009), implemented using Canty and Ripley (2014) and Hayfield and Racine (2008).”
Direct and indirect treatment effects–causal chains and mediation analysis with instrumental variables (2014, methodology): “This is equally attractive as directly manipulating the mediator exogenously; see the discussion in Imai et al. (2013). Perfect instruments are, however, rare in applications. Ten Have et al. (2007) also assumed treatment exogeneity but exploited treatment–covariate interactions as instruments for t”
Does Education Matter for Economic Growth? (2014, neutral): “[31] Hayfield, T.”
FROM THE BLAZAR SEQUENCE TO THE BLAZAR ENVELOPE: REVISITING THE RELATIVISTIC JET DICHOTOMY IN RADIO-LOUD ACTIVE GALACTIC NUCLEI (2011, methodology): “We utilized package np (Hayfield & Racine 2008) in the R language and environment for statistical computing (R Development Core Team 2010).”
Asymptotic normality and efficiency of two Sobol index estimators (2013, methodology): “ue model, and a Gaussian white noise ǫi of standard deviation 0.3 (yielding to a signal-to-noise ratio of 90%). The nonparametric regressions are carried using a Gaussian kernel (24), the R package np[7], together with the extrapolation method of [21] for window selection and the FIGtree [18] C++ li”
Optimal Bandwidth Selection for Nonparametric Conditional Distribution and Quantile Functions (2013, methodology): “We direct the interested reader to an implementation of (4) that is available in the R (R Core Team 2012) package “np” (see the function “npcdistbw,” version 0.40-14 and higher, Hayfield and Racine 2008).”
Including Covariates in the Regression Discontinuity Design (2018, methodology): “…of m+(x, z) and m−(x, z) (using local linear regression) as well as Kh(u) required in (13), (For m+(X,Z) and m−(X,Z), the bandwidth selector CV only uses treated and nontreated observations, respectively.) based on the np package for the statistical software R by Hayfield and Racine (2008).”
Most Likely Transformations (2015, methodology): “Hothorn et al. (2014) decoupled the parameterisation of the conditional transformation function h(y | x) from the estimation procedure, and showed that many interesting and novel models can be understood as transformation models.”