library(spsur)
library(spdep)
library(spatialreg)
library(sf)
The spsur R-package (Mínguez, López, and Mur 2022; López, Mı́nguez, and Mur 2020) has been designed for estimate spatial regression models in a multiequational framework. However, because of its flexibility, it is also possible to obtain useful results for uniequational models. On the oher hand, the spatialreg package (Bivand, Pebesma, and Gómez-Rubio (2013); Bivand and Piras (2015)), is the most adequate alternative for working with uniequational models, in a pure cross-sectional setting. The purpose of this vignette is to compare the results of spsur and spatialreg in case of uniequational models. As we will see, the differences between the two are negligible.
In sum, the purpose of this vignette is to compare the results of spsur and spatialreg in the following issues:
Throughout the vignette we use the dataset NCOVR (National Consortium on Violence Research). These data were employed by Baller et al. (2001) to analyze the incidence of homicides rates in the US counties. The dataset can be freely dowloaded from https://geodacenter.github.io/data-and-lab/ncovr/
NCOVR contains 3085 spatial units (counties), for 4 different cross-sections (1960, 1970, 1980, 1990) and 69 variables. According to Baller et al. (2001), the main variables of interest are:
First, we can read the NCOVR dataset as a simple feature (sf) object (named NCOVR.sf),
data(NCOVR, package = "spsur")
The first three observations in NCOVR appear below:
<- st_drop_geometry(NCOVR.sf)
NCOVR ::kable(
knitrhead((NCOVR[1:3, 1:6])),
caption = 'First observations of NCOVR dataset' )
NAME | STATE_NAME | FIPS | SOUTH | HR60 | HR70 |
---|---|---|---|---|---|
Lake of the Woods | Minnesota | 27077 | 0 | 0.000000 | 0.000000 |
Ferry | Washington | 53019 | 0 | 0.000000 | 0.000000 |
Stevens | Washington | 53065 | 0 | 1.863863 | 1.915158 |
Whereas the geometry of the USA counties is shown in Figure 1:
plot(st_geometry(NCOVR.sf))
Following Baller et al. (2001), we consider a W matrix based on the criterion of 10 nearest-neighbourhood, which is immediate to obtain using the spdep package (Bivand, Pebesma, and Gómez-Rubio (2013); Bivand and Wong (2018)). The resulting weighting matrix will be called listw. Note that this matrix is non-symmetric and it is row-standardized.
# Obtain coordinates of centroids
<- sf::st_coordinates(sf::st_centroid(NCOVR.sf))
co <- spdep::nb2listw(spdep::knn2nb(spdep::knearneigh(co, k = 10,longlat = TRUE))) listw
Baller et al. (2001) specify a single linear model to explain the case of HR in the year 1960 with the results shown in the Table below,
\[\begin{equation} HR_{60} = \beta_{0}+\beta_{1}RD_{60}+\beta_{2}PS_{60}+\beta_{3}MA_{60}+\beta_{4}DV_{60} +\beta_{5}UE_{60} +\beta_{6}SOUTH+\epsilon_{60} \tag{2.1} \end{equation}\]
<- HR60 ~ RD60 + PS60 + MA60 + DV60 + UE60 + SOUTH
formula_60 <- stats::lm(formula = formula_60, data = NCOVR.sf)
lm_60 summary(lm_60)
#>
#> Call:
#> stats::lm(formula = formula_60, data = NCOVR.sf)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -14.026 -2.217 -0.635 1.393 88.312
#>
#> Coefficients:
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept) 8.12591 0.63465 12.804 < 2e-16 ***
#> RD60 1.79824 0.12341 14.571 < 2e-16 ***
#> PS60 0.35871 0.09216 3.892 0.000101 ***
#> MA60 -0.23047 0.01932 -11.931 < 2e-16 ***
#> DV60 1.16002 0.09483 12.233 < 2e-16 ***
#> UE60 -0.06195 0.03515 -1.762 0.078138 .
#> SOUTH 2.63862 0.23325 11.312 < 2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#>
#> Residual standard error: 4.743 on 3078 degrees of freedom
#> Multiple R-squared: 0.2966, Adjusted R-squared: 0.2952
#> F-statistic: 216.3 on 6 and 3078 DF, p-value: < 2.2e-16
The model can be estimated in usual way with the function stats::lm
(R Core Team (2019)),
<- HR60 ~ RD60 + PS60 + MA60 + DV60 + UE60 + SOUTH
formula_60 <- lm(formula = formula_60, data = NCOVR.sf)
lm_60 summary(lm_60)
#>
#> Call:
#> lm(formula = formula_60, data = NCOVR.sf)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -14.026 -2.217 -0.635 1.393 88.312
#>
#> Coefficients:
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept) 8.12591 0.63465 12.804 < 2e-16 ***
#> RD60 1.79824 0.12341 14.571 < 2e-16 ***
#> PS60 0.35871 0.09216 3.892 0.000101 ***
#> MA60 -0.23047 0.01932 -11.931 < 2e-16 ***
#> DV60 1.16002 0.09483 12.233 < 2e-16 ***
#> UE60 -0.06195 0.03515 -1.762 0.078138 .
#> SOUTH 2.63862 0.23325 11.312 < 2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#>
#> Residual standard error: 4.743 on 3078 degrees of freedom
#> Multiple R-squared: 0.2966, Adjusted R-squared: 0.2952
#> F-statistic: 216.3 on 6 and 3078 DF, p-value: < 2.2e-16
The same results can be obtained with the function spsurml
from spsur, selecting the argument type = “sim”
<- spsurml(formula = formula_60, type = "sim",
ols.spsur data = NCOVR.sf)
#> Initial point:
#> log_lik: -9176.338
#> Iteration: 1 log_lik: -9176.338
#> Time to fit the model: 0.07 seconds
#> Time to compute covariances: 0 seconds
summary(ols.spsur)
#> Call:
#> spsurml(formula = formula_60, data = NCOVR.sf, type = "sim")
#>
#>
#> Spatial SUR model type: sim
#>
#> Equation 1
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept)_1 8.125915 0.634033 12.8162 < 2.2e-16 ***
#> RD60_1 1.798240 0.123294 14.5850 < 2.2e-16 ***
#> PS60_1 0.358706 0.092069 3.8960 9.986e-05 ***
#> MA60_1 -0.230475 0.019298 -11.9428 < 2.2e-16 ***
#> DV60_1 1.160020 0.094737 12.2446 < 2.2e-16 ***
#> UE60_1 -0.061948 0.035120 -1.7639 0.07785 .
#> SOUTH_1 2.638618 0.233024 11.3234 < 2.2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> R-squared: 0.2966
#>
#> Residual standard error: 4.739
The two functions, stats::lm
and spsurml
, produce identical results. The output of spsurml
is shorter than that of lm
so, depending on the necessities of the user, he/she can choose between lm
and spsurml
without loss of information. If you want all the estimation details, then use lm
.
The existence of omitted spatial dependence in the results of a SIM model, estimated by LS, can be tested by using the classical LM tests. The function spdep::lm.LMtests
report the values of these Lagrange Multipliers
<- spdep::lm.LMtests(lm_60, listw, test = "all")
lmtest.spdep print(lmtest.spdep)
#>
#> Lagrange multiplier diagnostics for spatial dependence
#>
#> data:
#> model: lm(formula = formula_60, data = NCOVR.sf)
#> weights: listw
#>
#> LMerr = 208.12, df = 1, p-value < 2.2e-16
#>
#>
#> Lagrange multiplier diagnostics for spatial dependence
#>
#> data:
#> model: lm(formula = formula_60, data = NCOVR.sf)
#> weights: listw
#>
#> LMlag = 232.1, df = 1, p-value < 2.2e-16
#>
#>
#> Lagrange multiplier diagnostics for spatial dependence
#>
#> data:
#> model: lm(formula = formula_60, data = NCOVR.sf)
#> weights: listw
#>
#> RLMerr = 1.1565, df = 1, p-value = 0.2822
#>
#>
#> Lagrange multiplier diagnostics for spatial dependence
#>
#> data:
#> model: lm(formula = formula_60, data = NCOVR.sf)
#> weights: listw
#>
#> RLMlag = 25.138, df = 1, p-value = 5.337e-07
#>
#>
#> Lagrange multiplier diagnostics for spatial dependence
#>
#> data:
#> model: lm(formula = formula_60, data = NCOVR.sf)
#> weights: listw
#>
#> SARMA = 233.25, df = 2, p-value < 2.2e-16
The same tests can be obtained with the function lmtestspsur
<- lmtestspsur(formula = formula_60, listw = listw,
lmtest.spsur data = NCOVR.sf)
print(lmtest.spsur)
#> [[1]]
#>
#> LM-SUR-SLM
#>
#> data: NCOVR.sf
#> LM-stat = 231.96, df = 1, p-value < 2.2e-16
#>
#>
#> [[2]]
#>
#> LM-SUR-SEM
#>
#> data: NCOVR.sf
#> LM-stat = 207.98, df = 1, p-value < 2.2e-16
#>
#>
#> [[3]]
#>
#> LM*-SUR-SLM
#>
#> data: NCOVR.sf
#> LM-stat = 25.13, df = 1, p-value = 5.36e-07
#>
#>
#> [[4]]
#>
#> LM*-SUR-SEM
#>
#> data: NCOVR.sf
#> LM-stat = 1.1523, df = 1, p-value = 0.2831
#>
#>
#> [[5]]
#>
#> LM-SUR-SARAR
#>
#> data: NCOVR.sf
#> LM-stat = 233.11, df = 2, p-value < 2.2e-16
Note that the ordering of the battery of Lagrange Multipliers is not the same. Otherwise, the results are almost identical.
Both R packages spatialreg and spsur can estimate Spatial Lag Models for a single cross-section. Continuing with the example before, the model that we want to estimate is:
\[ \begin{equation} HR_{60} = \rho W HR_{60} +\beta_{0}+\beta_{1}RD_{60}+\beta_{2}PS_{60}+\beta_{3}MA_{60}+\beta_{4}DV_{60} +\beta_{5}UE_{60} +\beta_{6}SOUTH+\epsilon_{60}\ \tag{4.1} \end{equation} \]
The ML estimation of equation (4.1) using the function spsurml()
of spsur renders the following results:
<- spsur::spsurml(formula = formula_60, type = "slm",
slm.spsur listw = listw, data = NCOVR.sf)
#> neighbourhood matrix eigenvalues
#> Computing eigenvalues ...
#>
#> Initial point: log_lik: -9101.721 rhos: 0.364
#> Iteration: 1 log_lik: -9098.57 rhos: 0.37
#> Iteration: 2 log_lik: -9098.569 rhos: 0.37
#> Time to fit the model: 1.18 seconds
#> Time to compute covariances: 5.68 seconds
summary(slm.spsur)
#> Call:
#> spsur::spsurml(formula = formula_60, data = NCOVR.sf, listw = listw,
#> type = "slm")
#>
#>
#> Spatial SUR model type: slm
#>
#> Equation 1
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept)_1 5.475515 0.647627 8.4547 < 2.2e-16 ***
#> RD60_1 1.358071 0.125738 10.8008 < 2.2e-16 ***
#> PS60_1 0.284375 0.089320 3.1838 0.001468 **
#> MA60_1 -0.165479 0.019384 -8.5367 < 2.2e-16 ***
#> DV60_1 0.884739 0.093725 9.4398 < 2.2e-16 ***
#> UE60_1 -0.021741 0.034004 -0.6394 0.522641
#> SOUTH_1 1.355704 0.244626 5.5419 3.244e-08 ***
#> rho_1 0.370414 0.030275 12.2350 < 2.2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> R-squared: 0.3409
#>
#> Residual standard error: 4.588
#> LMM: 22.393 p-value: (2.22e-06)
The output of the function spatialreg::lagsarlm()
from spatialreg is
<- spatialreg::lagsarlm(formula = formula_60,
slm.spatialreg listw = listw, type = "lag",
data = NCOVR.sf)
summary(slm.spatialreg)
#>
#> Call:
#> spatialreg::lagsarlm(formula = formula_60, data = NCOVR.sf, listw = listw,
#> type = "lag")
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -13.54416 -2.11106 -0.63828 1.30754 88.57312
#>
#> Type: lag
#> Coefficients: (asymptotic standard errors)
#> Estimate Std. Error z value Pr(>|z|)
#> (Intercept) 5.475272 0.647529 8.4556 < 2.2e-16
#> RD60 1.358031 0.125719 10.8021 < 2.2e-16
#> PS60 0.284368 0.089306 3.1842 0.001452
#> MA60 -0.165474 0.019381 -8.5377 < 2.2e-16
#> DV60 0.884714 0.093710 9.4410 < 2.2e-16
#> UE60 -0.021737 0.033999 -0.6393 0.522596
#> SOUTH 1.355586 0.244590 5.5423 2.986e-08
#>
#> Rho: 0.37045, LR test value: 155.54, p-value: < 2.22e-16
#> Asymptotic standard error: 0.030274
#> z-value: 12.237, p-value: < 2.22e-16
#> Wald statistic: 149.74, p-value: < 2.22e-16
#>
#> Log likelihood: -9098.569 for lag model
#> ML residual variance (sigma squared): 21.043, (sigma: 4.5872)
#> Number of observations: 3085
#> Number of parameters estimated: 9
#> AIC: 18215, (AIC for lm: 18369)
#> LM test for residual autocorrelation
#> test value: 22.435, p-value: 2.1742e-06
Once again, the two estimations are almost the same. The estimated log-likelihood of the SLM can be recovered using the function logLik()
logLik(slm.spsur)
#> 'log Lik.' -9098.569 (df=9)
The log-likelihood corresponding to the SIM model is much lower, -9176.338, which points to a severe misspecification in the last model. More formally, the LR, obtained with the function anova
of spsur strongly rejects the SIM model in favour of the SLM alternative; the AIC and BIC statistics indicates the same.
anova(ols.spsur, slm.spsur)
#> logLik df AIC BIC LRtest p.val
#> model 1: sim -9176.3 8 18369 18353
#> model 2: slm -9098.6 9 18215 18197 155.54 1.0684e-35
The SLM model can also be estimate by Three Stages-Least-Squares (3SLS) by both R-packages. The function for spsur is spsur3sls()
.3sls.spsur <- spsur3sls(formula = formula_60, type = "slm",
slmlistw = listw, data = NCOVR.sf)
#> Time to fit the model: 0.05 seconds
summary(slm.3sls.spsur)
#> Call:
#> spsur3sls(formula = formula_60, data = NCOVR.sf, listw = listw,
#> type = "slm")
#>
#>
#> Spatial SUR model type: slm
#>
#> Equation 1
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept)_1 4.00373231 0.83096624 4.8182 1.519e-06 ***
#> RD60_1 1.11364326 0.15193264 7.3298 2.931e-13 ***
#> PS60_1 0.24309853 0.09249060 2.6284 0.008622 **
#> MA60_1 -0.12938710 0.02331379 -5.5498 3.103e-08 ***
#> DV60_1 0.73187418 0.10955683 6.6803 2.819e-11 ***
#> UE60_1 0.00058645 0.03576227 0.0164 0.986917
#> SOUTH_1 0.64329382 0.35018066 1.8370 0.066301 .
#> rho_1 0.57610680 0.07601797 7.5786 4.595e-14 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> R-squared: 0.3447
#>
#> Residual standard error: 4.695
Whereas spatialreg uses the function spatialreg::stsls()
.3sls.spatialreg <- spatialreg::stsls(formula = formula_60, listw = listw,
slmdata = NCOVR.sf)
summary(slm.3sls.spatialreg)
#>
#> Call:
#> spatialreg::stsls(formula = formula_60, data = NCOVR.sf, listw = listw)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -14.78215 -2.06681 -0.57751 1.22823 88.71807
#>
#> Coefficients:
#> Estimate Std. Error t value Pr(>|t|)
#> Rho 0.57610680 0.07413902 7.7706 7.772e-15
#> (Intercept) 4.00373231 0.81042715 4.9403 7.801e-07
#> RD60 1.11364326 0.14817730 7.5156 5.662e-14
#> PS60 0.24309853 0.09020450 2.6950 0.007039
#> MA60 -0.12938710 0.02273754 -5.6905 1.267e-08
#> DV60 0.73187418 0.10684890 6.8496 7.405e-12
#> UE60 0.00058645 0.03487833 0.0168 0.986585
#> SOUTH 0.64329382 0.34152520 1.8836 0.059620
#>
#> Residual variance (sigma squared): 20.967, (sigma: 4.579)
There is hardly any difference between them because the estimation algorithm is quasilinear and both functions use the same set of instruments. The case of the SDM model produces identical results.
The model to estimate in this case is:
\[ \begin{equation} HR_{60} = \beta_{0}+\beta_{1}RD_{60}+\beta_{2}PS_{60}+\beta_{3}MA_{60}+\beta_{4}DV_{60} +\beta_{5}UE_{60} +\beta_{6}SOUTH+ u_{60} \\ u_{60} = \lambda W u_{60} + \epsilon_{60}\ \tag{5.1} \end{equation} \]
which can be solved by ML using the function spatialreg::errorsarlm()
, from spatialreg.
<- spatialreg::errorsarlm(formula = formula_60,
sem.spatialreg listw = listw, data = NCOVR.sf)
summary(sem.spatialreg)
#>
#> Call:spatialreg::errorsarlm(formula = formula_60, data = NCOVR.sf,
#> listw = listw)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -13.54014 -2.10753 -0.66838 1.27875 88.31401
#>
#> Type: error
#> Coefficients: (asymptotic standard errors)
#> Estimate Std. Error z value Pr(>|z|)
#> (Intercept) 7.527982 0.766617 9.8197 < 2.2e-16
#> RD60 1.742357 0.148632 11.7226 < 2.2e-16
#> PS60 0.346866 0.109221 3.1758 0.001494
#> MA60 -0.208285 0.023388 -8.9057 < 2.2e-16
#> DV60 0.957248 0.108389 8.8316 < 2.2e-16
#> UE60 0.018210 0.039933 0.4560 0.648383
#> SOUTH 2.495326 0.316229 7.8909 3.109e-15
#>
#> Lambda: 0.38244, LR test value: 140.48, p-value: < 2.22e-16
#> Asymptotic standard error: 0.032062
#> z-value: 11.928, p-value: < 2.22e-16
#> Wald statistic: 142.28, p-value: < 2.22e-16
#>
#> Log likelihood: -9106.1 for error model
#> ML residual variance (sigma squared): 21.124, (sigma: 4.5961)
#> Number of observations: 3085
#> Number of parameters estimated: 9
#> AIC: 18230, (AIC for lm: 18369)
spsur always uses the same function for the ML algorithm, spsurml()
; you only have to change the type of model to estimate.
<- spsurml(formula = formula_60, type = "sem",
sem.spsur listw = listw, data = NCOVR.sf)
#> neighbourhood matrix eigenvalues
#> Computing eigenvalues ...
#>
#> Initial point: log_lik: -9108.886 lambdas: 0.375
#> Iteration: 1 log_lik: -9106.101 lambdas: 0.382
#> Iteration: 2 log_lik: -9106.1 lambdas: 0.382
#> Time to fit the model: 2.52 seconds
#> Time to compute covariances: 97.7 seconds
summary(sem.spsur)
#> Call:
#> spsurml(formula = formula_60, data = NCOVR.sf, listw = listw,
#> type = "sem")
#>
#>
#> Spatial SUR model type: sem
#>
#> Equation 1
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept)_1 7.528052 0.766724 9.8185 < 2.2e-16 ***
#> RD60_1 1.742363 0.148653 11.7210 < 2.2e-16 ***
#> PS60_1 0.346864 0.109237 3.1753 0.001511 **
#> MA60_1 -0.208288 0.023391 -8.9046 < 2.2e-16 ***
#> DV60_1 0.957272 0.108405 8.8305 < 2.2e-16 ***
#> UE60_1 0.018201 0.039939 0.4557 0.648628
#> SOUTH_1 2.495357 0.316267 7.8900 4.163e-15 ***
#> lambda_1 0.382398 0.032109 11.9095 < 2.2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> R-squared: 0.3387
#>
#> Residual standard error: 4.597
#> LMM: 1.193 p-value: (0.275)
The LR test for nested models, in this case SIM vs SEM, can be obtained as before using the function anova()
anova(ols.spsur, sem.spsur)
#> logLik df AIC BIC LRtest p.val
#> model 1: sim -9176.3 8 18369 18353
#> model 2: sem -9106.1 9 18230 18212 140.48 2.0951e-32
Clearly, the SEM model is preferable to the SIM specification because there is spatial dependence in the data, which has been effectively captured by the SEM mechanism. Moreover, according to the LMM test below (this is a Marginal Lagrange Multiplier of the type \(LM(\rho|\lambda)\); see the spsur user guide), once the spatial errors are introduced in the equation, the spatial lag of the endogenous variable, \(WHR_{60}\) is not statistically significant.
print(sem.spsur$LMM)
#> [1] 1.193044
The especification of the SARAR model is as usual
\[ \begin{equation} HR_{60} = \rho W HR_{60} + \beta_{0}+\beta_{1}RD_{60}+\beta_{2}PS_{60}+\beta_{3}MA_{60}+\beta_{4}DV_{60} +\beta_{5}UE_{60} +\beta_{6}SOUTH+ u_{60} \\ u_{60} = \lambda W u_{60} + \epsilon_{60}\ \tag{6.1} \end{equation} \]
spsur estimates this model by using the funtion spsurml()
; you only have to adjust the argument type to sarar, that is
<- spsurml(formula = formula_60, listw = listw,
sarar.spsur type ="sarar",data = NCOVR.sf)
#> neighbourhood matrix eigenvalues
#> Computing eigenvalues ...
#>
#> Initial point: log_lik: -9085.419 rhos: 0.663 lambdas: -0.603
#> Iteration: 1 log_lik: -9062.889 rhos: 0.748 lambdas: -0.837
#> Iteration: 2 log_lik: -9060.502 rhos: 0.769 lambdas: -0.905
#> Iteration: 3 log_lik: -9060.28 rhos: 0.775 lambdas: -0.925
#> Iteration: 4 log_lik: -9060.259 rhos: 0.777 lambdas: -0.93
#> Iteration: 5 log_lik: -9060.257 rhos: 0.777 lambdas: -0.932
#> Iteration: 6 log_lik: -9060.256 rhos: 0.778 lambdas: -0.932
#> Time to fit the model: 46.67 seconds
#> Time to compute covariances: 9.31 seconds
summary(sarar.spsur)
#> Call:
#> spsurml(formula = formula_60, data = NCOVR.sf, listw = listw,
#> type = "sarar")
#>
#>
#> Spatial SUR model type: sarar
#>
#> Equation 1
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept)_1 2.424897 0.391608 6.1922 6.723e-10 ***
#> RD60_1 0.657738 0.079705 8.2521 2.281e-16 ***
#> PS60_1 0.149831 0.054133 2.7678 0.005677 **
#> MA60_1 -0.079625 0.011841 -6.7245 2.093e-11 ***
#> DV60_1 0.500529 0.063120 7.9298 3.044e-15 ***
#> UE60_1 -0.031844 0.021379 -1.4895 0.136455
#> SOUTH_1 0.261410 0.127913 2.0437 0.041072 *
#> rho_1 0.777641 0.018131 42.8893 < 2.2e-16 ***
#> lambda_1 -0.932459 0.083230 -11.2034 < 2.2e-16 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> R-squared: 0.444
#>
#> Residual standard error: 4.241
In the case of spatialreg, the function for this case is spatialreg::sacsar()
<- spatialreg::sacsarlm(formula = formula_60,
sarar.spatialreg listw = listw, data = NCOVR.sf)
summary(sarar.spatialreg)
#>
#> Call:
#> spatialreg::sacsarlm(formula = formula_60, data = NCOVR.sf, listw = listw)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -12.72590 -1.99405 -0.57621 1.20291 80.72854
#>
#> Type: sac
#> Coefficients: (asymptotic standard errors)
#> Estimate Std. Error z value Pr(>|z|)
#> (Intercept) 2.422847 0.410239 5.9059 3.506e-09
#> RD60 0.657260 0.085075 7.7256 1.110e-14
#> PS60 0.149719 0.054436 2.7504 0.005953
#> MA60 -0.079565 0.012404 -6.4146 1.412e-10
#> DV60 0.500200 0.066423 7.5305 5.063e-14
#> UE60 -0.031823 0.021382 -1.4883 0.136671
#> SOUTH 0.260793 0.131656 1.9809 0.047606
#>
#> Rho: 0.77788
#> Asymptotic standard error: 0.023069
#> z-value: 33.72, p-value: < 2.22e-16
#> Lambda: -0.93327
#> Asymptotic standard error: 0.072704
#> z-value: -12.837, p-value: < 2.22e-16
#>
#> LR test value: 232.16, p-value: < 2.22e-16
#>
#> Log likelihood: -9060.256 for sac model
#> ML residual variance (sigma squared): 17.974, (sigma: 4.2396)
#> Number of observations: 3085
#> Number of parameters estimated: 10
#> AIC: 18141, (AIC for lm: 18369)
The same as before, the differences between the two codes are minimal.
Finally, according to the anova()
function of spsur, the SARAR model is preferable to the SLM and SEM alternatives, which are nested in the SARAR, in spite of the high value estimated for the parameter \(\lambda\), -0.93327, close to a unit root case. The LR and both AIC and BIC statistics reach the same conclusion.
anova(slm.spsur,sarar.spsur)
#> logLik df AIC BIC LRtest p.val
#> model 1: slm -9098.6 9 18215 18197
#> model 2: sarar -9060.3 10 18141 18121 76.625 2.0666e-18
anova(sem.spsur,sarar.spsur)
#> logLik df AIC BIC LRtest p.val
#> model 1: sem -9106.1 9 18230 18212
#> model 2: sarar -9060.3 10 18141 18121 91.687 1.0151e-21