2

我需要单独估计空间滞后 X (SLX) 的空间计量经济学模型,结合空间自回归模型 (SAR) 或空间误差模型 (SEM)。根据 Vega & Elhorst (2015) 的论文“The SLX Model”,将它们组合起来后,它们被称为空间杜宾模型 (SDM) 或空间杜宾误差模型 (SDEM)。

我打算使用 splm 包估计 R 中的所有空间面板模型,这也需要 spdep 函数。从这个意义上说,我从一个形状文件创建了 Queen 和 k = 4 类型的邻居列表:

> TCAL <- readOGR(dsn = ".", "Municipios_csv")
> coords <- coordinates(TCAL)

> contnbQueen <- poly2nb(TCAL, queen = TRUE)
> enter code herecontnbk4    <- knn2nb(knearneigh(coords, k = 4, RANN = FALSE))

然后我将此邻居列表转换为权重矩阵:

> W <- nb2listw(contnbk4, glist = NULL, style = "W")

attributes(W)
$names
[1] "style"      "neighbours" "weights"   

$class
[1] "listw" "nb"   

$region.id
 [1] "1"   "2"   "3"   "4"   "5"   "6"   "7"   "8"   "9"   "10"  "11"  "12"  "13"  "14"  "15"  "16"  "17"  "18"  "19"  "20" 
 [21] "21"  "22"  "23"  "24"  "25"  "26"  "27"  "28"  "29"  "30"  "31"  "32"  "33"  "34"  "35"  "36"  "37"  "38"  "39"  "40" 
 [41] "41"  "42"  "43"  "44"  "45"  "46"  "47"  "48"  "49"  "50"  "51"  "52"  "53"  "54"  "55"  "56"  "57"  "58"  "59"  "60" 
 [61] "61"  "62"  "63"  "64"  "65"  "66"  "67"  "68"  "69"  "70"  "71"  "72"  "73"  "74"  "75"  "76"  "77"  "78"  "79"  "80" 
 [81] "81"  "82"  "83"  "84"  "85"  "86"  "87"  "88"  "89"  "90"  "91"  "92"  "93"  "94"  "95"  "96"  "97"  "98"  "99"  "100"
[101] "101" "102" "103" "104" "105" "106" "107" "108" "109" "110" "111" "112" "113" "114" "115" "116" "117" "118" "119" "120"
[121] "121" "122" "123" "124" "125" "126" "127" "128" "129" "130" "131" "132" "133" "134" "135" "136" "137" "138" "139" "140"
[141] "141" "142" "143" "144" "145" "146" "147" "148" "149" "150" "151" "152" "153" "154" "155" "156" "157" "158" "159" "160"
[161] "161" "162" "163" "164" "165" "166" "167" "168" "169" "170" "171" "172" "173" "174" "175" "176" "177" "178" "179" "180"
[181] "181" "182" "183" "184" "185" "186" "187" "188" "189" "190" "191" "192" "193" "194" "195" "196" "197" "198" "199" "200"
[201] "201" "202" "203" "204" "205" "206" "207" "208" "209" "210" "211" "212" "213" "214" "215" "216" "217" "218" "219" "220"
[221] "221" "222" "223" "224" "225" "226" "227" "228" "229" "230" "231" "232" "233" "234" "235" "236" "237" "238" "239" "240"
[241] "241" "242" "243" "244" "245" "246" "247" "248" "249" "250" "251" "252" "253" "254" "255" "256" "257" "258" "259" "260"
[261] "261" "262" "263" "264" "265" "266" "267" "268" "269" "270" "271" "272" "273" "274" "275" "276"

    $call
    nb2listw(neighbours = contnbk7, glist = NULL, style = "W")

下一步,我为面板 SAR 和 SEM 模型创建了一个公式,它运行良好并产生了估计值:

> fmPanel <- Area ~ Dist + Land + CredAg
> vegSAR <- spml(fmPanel, data = veg, index = c("Mun","Year"), listw = W, model = "within", effect = "twoways", spatial.error = "none", lag = TRUE)
> vegSEM <- spml(fmPanel, data = veg, index = c("Mun","Year"), listw = W, model = "within", effect = "twoways", spatial.error = "b", lag = FALSE)

然后,我尝试通过创建协变量 X 的空间滞后来估计 SLX、SDM 和 SDEM 模型:

> vegX <- pdata.frame(veg, index = c("Mun","Year")); class(vegX)

    [1] "pdata.frame" "data.frame"

然后我创建了 pseries 值:

> DistX <- vegX$Dist; class(DistX)
    [1] "pseries" "numeric"
> LandX <- vegX$Land; class(LandX)
    [1] "pseries" "numeric"
> CredAgX <- vegX$CredAg; class(CredAgX)
    [1] "pseries" "numeric"

但是当我应用slag函数时发生了错误:

DistX <- slag(agSPX$Dist, listw = W)

    Error in lag.listw(listw, xt) : object lengths differ

我的面板数据有 5 年和 276 个区域。因此,对象的特征是:

> length(DistX)
[1] 1380

> length(W)
[1] 3

> length(W$weights)
[1] 276

所以,我想知道,如果我可以在矩阵中转换 W$weights,例如用作slag 函数示例的usaww,我可以应用函数mat2listw,然后在 X 上使用 slag

有人能告诉我我哪里错了吗?

4

2 回答 2

2

可能这不是最好的解决方案,但我通过以下步骤计算了 SLX、SDM 和 SDEM 模型:

1)通过以下方式加载形状文件:

TCAL <- readOGR(dsn = ".", "Municipios_csv_BIO")
coords <- coordinates(TCAL)

2)通过以下方式创建加权矩阵 W:

contnbQueen  <- poly2nb(TCAL, queen = TRUE)
contnbk4     <- knn2nb(knearneigh(coords, k = 4,  RANN = FALSE))

3)选择要应用的矩阵:

W <- nb2listw(contnbk4, glist = NULL, style = "W")

4)在pdata.frame中转换data.frame:

vegSPX  <- pdata.frame(vegPainel, index = c("ID","Ano"))

5)创建特定的 pdata.frame 空间向量,例如:

vegIDDX  <- vegSPX$IDD
vegSoilX <- vegSPX$Soil
vegQAIX  <- vegSPX$QAI

6)为这些模型指定一个公式:

fmSPvegX <- Area ~ IDD + Soil + QAI + slag(vegIDDX, listw = W) + slag(vegSoilX, listw = W) + slag(vegQAIX, listw = W)

7)对原始数据(data.table data.frame对象)应用plm和splm函数:

vegSLX<-plm(fmSPvegX,data=vegPainel,listw=W,index=c("ID","Ano"),model="within",effect="twoways",spatial.error="none",lag=F)
vegSARX<-spml(fmSPvegX,data=vegPainel,listw=W,index=c("ID","Ano"),model="within",effect="twoways",spatial.error="none",lag=T)
vegSEMX<-spml(fmSPvegX,data=vegPainel,listw=W,index=c("ID","Ano"),model="within",effect="twoways",spatial.error="b",lag=F)

我希望它会有所帮助。

于 2017-12-29T13:36:51.240 回答
1

您还可以将滞后的解释变量添加到数据中。

一个可重现的例子:

library(plm)
library(spatialreg)
library(splm)

# load data
data(Produc, package = "plm")
data(usaww, package = "splm")

d <- pdata.frame(Produc, index = c("state","year"), drop.index = FALSE)

# create a spatial explanatory variable
d$unemp_l <- slag(d$unemp, usaww)

# run model
m <- splm::spml(log(gsp) ~ log(pcap) + log(pc) + log(emp) + unemp + unemp_l,
                  data = d, listw = mat2listw(usaww) , model="within")
summary(m)
Spatial panel fixed effects error model


Call:
splm::spml(formula = log(gsp) ~ log(pcap) + log(pc) + log(emp) + 
    unemp + unemp_l, data = d, listw = mat2listw(usaww), model = "within")

Residuals:
      Min.    1st Qu.     Median    3rd Qu.       Max. 
-0.1211492 -0.0234013 -0.0040218  0.0167919  0.1787587 

Spatial error parameter:
    Estimate Std. Error t-value  Pr(>|t|)    
rho 0.542254   0.033772  16.056 < 2.2e-16 ***

Coefficients:
            Estimate Std. Error t-value Pr(>|t|)    
log(pcap)  0.0090575  0.0251036  0.3608  0.71824    
log(pc)    0.2152367  0.0234077  9.1951  < 2e-16 ***
log(emp)   0.7833003  0.0277672 28.2096  < 2e-16 ***
unemp     -0.0014795  0.0011443 -1.2930  0.19603    
unemp_l   -0.0031210  0.0015790 -1.9766  0.04808 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

您可以验证空间解释是否正确。例如:

a <- usaww["ALABAMA",]
a <- a[a!=0]
a
    FLORIDA     GEORGIA MISSISSIPPI    TENNESSE 
       0.25        0.25        0.25        0.25 
mean(d[d$year=="1970" & d$state %in% names(a) , "unemp"])
[1] 4.525

d[d$state=="ALABAMA" & d$year=="1970", "unemp_l"]
ALABAMA-1970 
   4.525 
于 2020-04-28T09:13:58.977 回答