project1<-read.csv("C:/Users/Allan/OneDrive/Desktop/taluremia  clean data  analysis 1.csv")
project1<-na.omit(project1)
#selecting data for NE region
NE_region<-project1[project1$county=="NE",]
# modelling for NE region
library(plm)
names(project1)
##  [1] "county"          "state"           "state.1"         "year"           
##  [5] "year.1"          "forest"          "mixeddev"        "CumTemp"        
##  [9] "CumPrecip"       "TempVar"         "PrecipVar"       "AvgWinterTemp"  
## [13] "AvgSpringPrecip" "HotDryDays"      "talremia_inc"    "tick_lag"
NE_region_model1<-plm(talremia_inc~TempVar+tick_lag,data=NE_region,index=c("county","year.1"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(NE_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ TempVar + tick_lag, data = NE_region, 
##     model = "within", index = c("county", "year.1"))
## 
## Balanced Panel: n = 1, T = 10, N = 110
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -2.96859 -1.08948 -0.33076  0.54145  7.68691 
## 
## Coefficients:
##            Estimate Std. Error t-value  Pr(>|t|)    
## TempVar  -0.0143082  0.0028091 -5.0935 1.519e-06 ***
## tick_lag  0.0289948  0.0178801  1.6216    0.1078    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    419.42
## Residual Sum of Squares: 333.63
## R-Squared:      0.20454
## Adj. R-Squared: 0.18967
## F-statistic: 13.7568 on 2 and 107 DF, p-value: 4.82e-06
NE_region_model2<-plm(talremia_inc ~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays
                      +tick_lag,data=NE_region,index=c("county","year.1"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(NE_region_model2)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = NE_region, model = "within", index = c("county", 
##     "year.1"))
## 
## Balanced Panel: n = 1, T = 10, N = 110
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -2.55381 -0.93210 -0.20483  0.62293  5.33474 
## 
## Coefficients:
##                   Estimate Std. Error t-value  Pr(>|t|)    
## CumPrecip        0.0031428  0.0012749  2.4652  0.015398 *  
## forest          -0.0400331  0.0075991 -5.2682 7.935e-07 ***
## mixeddev        -1.8363538  0.4522750 -4.0603 9.749e-05 ***
## TempVar         -0.0027142  0.0053809 -0.5044  0.615084    
## PrecipVar       -0.0576219  0.0173020 -3.3304  0.001216 ** 
## AvgWinterTemp    0.0310621  0.0498556  0.6230  0.534676    
## AvgSpringPrecip -0.2959025  0.1281206 -2.3096  0.022969 *  
## HotDryDays      -0.0172457  0.0506623 -0.3404  0.734265    
## tick_lag         0.0442506  0.0168704  2.6230  0.010080 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    419.42
## Residual Sum of Squares: 202.33
## R-Squared:      0.51759
## Adj. R-Squared: 0.47417
## F-statistic: 11.9213 on 9 and 100 DF, p-value: 1.3975e-12
#cluster robust se
library(sandwich)
clustered_se <- vcovHC(NE_region_model2, type = "HC1", cluster = "group")

coef(NE_region_model2,vcov=clustered_se)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##     0.003142804    -0.040033130    -1.836353849    -0.002714184    -0.057621907 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##     0.031062055    -0.295902452    -0.017245721     0.044250634
# Midwest region
MW_region<-project1[project1$county=="MW",]
#modelling for midwest
MW_region_model1<-plm(talremia_inc~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays+tick_lag,data=MW_region,index=c("county","year"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(MW_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = MW_region, model = "within", index = c("county", 
##     "year"))
## 
## Balanced Panel: n = 1, T = 10, N = 120
## 
## Residuals:
##      Min.   1st Qu.    Median   3rd Qu.      Max. 
## -15.31231  -4.54494  -0.79275   2.99295  21.42614 
## 
## Coefficients:
##                    Estimate  Std. Error t-value  Pr(>|t|)    
## CumPrecip        -0.0325670   0.0061129 -5.3276 5.338e-07 ***
## forest            0.1275292   0.0624949  2.0406  0.043682 *  
## mixeddev        -10.4946630   3.7744853 -2.7804  0.006388 ** 
## TempVar          -0.0059059   0.0161253 -0.3663  0.714882    
## PrecipVar         0.4385078   0.0906899  4.8352 4.349e-06 ***
## AvgWinterTemp    -0.1376964   0.2308852 -0.5964  0.552144    
## AvgSpringPrecip   0.2088590   0.6059839  0.3447  0.731007    
## HotDryDays        0.5832773   0.2588939  2.2530  0.026245 *  
## tick_lag          0.1459472   0.0773339  1.8872  0.061765 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    10097
## Residual Sum of Squares: 5714.3
## R-Squared:      0.43405
## Adj. R-Squared: 0.38774
## F-statistic: 9.37364 on 9 and 110 DF, p-value: 1.8054e-10
#cluster robust se
library(sandwich)
clustered_se2 <- vcovHC(MW_region_model1, type = "HC1", cluster = "group")
coef(MW_region_model1,vcov=clustered_se2)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##    -0.032567047     0.127529166   -10.494662990    -0.005905907     0.438507815 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##    -0.137696400     0.208859043     0.583277290     0.145947197
#for southwest region
SW_region<-project1[project1$county=="SW",]
#modelling for south west region
SW_region_model1<-plm(talremia_inc~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays+tick_lag,data=SW_region,index=c("county","year"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(SW_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = SW_region, model = "within", index = c("county", 
##     "year"))
## 
## Balanced Panel: n = 1, T = 10, N = 40
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -14.4874  -4.7439  -2.1853   3.4764  42.0687 
## 
## Coefficients:
##                   Estimate Std. Error t-value Pr(>|t|)  
## CumPrecip        0.0356011  0.0270031  1.3184  0.19735  
## forest          -0.0016152  0.2574322 -0.0063  0.99504  
## mixeddev        -7.7419448  3.9550375 -1.9575  0.05965 .
## TempVar          0.0185897  0.0678414  0.2740  0.78595  
## PrecipVar       -0.4546685  0.3077552 -1.4774  0.15000  
## AvgWinterTemp    0.6230089  0.7871156  0.7915  0.43486  
## AvgSpringPrecip  0.7133847  2.7728912  0.2573  0.79873  
## HotDryDays       0.1619643  0.3603876  0.4494  0.65636  
## tick_lag         0.3604576  0.3205911  1.1244  0.26978  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    6363.6
## Residual Sum of Squares: 3473.2
## R-Squared:      0.4542
## Adj. R-Squared: 0.29047
## F-statistic: 2.77396 on 9 and 30 DF, p-value: 0.017186
#cluster robust se
library(sandwich)
clustered_se3 <- vcovHC(SW_region_model1, type = "HC1", cluster = "group")
coef(SW_region_model1,vcov=clustered_se3)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##     0.035601074    -0.001615153    -7.741944812     0.018589662    -0.454668536 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##     0.623008858     0.713384724     0.161964296     0.360457623
#pacific region
pa_region<-project1[project1$county=="P",]
#modelling for pacific region
pa_region_model1<-plm(talremia_inc~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays+tick_lag,data=pa_region,index=c("county","year"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(pa_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = pa_region, model = "within", index = c("county", 
##     "year"))
## 
## Balanced Panel: n = 1, T = 10, N = 70
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -5.62054 -1.14415 -0.28157  1.04382 10.96404 
## 
## Coefficients:
##                   Estimate Std. Error t-value  Pr(>|t|)    
## CumPrecip       -0.0046836  0.0023787 -1.9689   0.05358 .  
## forest           0.0477614  0.0653600  0.7307   0.46778    
## mixeddev        -0.9603335  7.3031866 -0.1315   0.89582    
## TempVar         -0.0041440  0.0097092 -0.4268   0.67104    
## PrecipVar        0.0588792  0.0405487  1.4521   0.15169    
## AvgWinterTemp   -0.0379396  0.0461967 -0.8213   0.41475    
## AvgSpringPrecip  0.7882544  0.3716843  2.1208   0.03808 *  
## HotDryDays       0.4970523  0.2363859  2.1027   0.03969 *  
## tick_lag         0.0668377  0.0129537  5.1597 2.939e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    653.37
## Residual Sum of Squares: 320.93
## R-Squared:      0.50881
## Adj. R-Squared: 0.43513
## F-statistic: 6.90574 on 9 and 60 DF, p-value: 9.3819e-07
#cluster robust se
library(sandwich)
clustered_se4 <- vcovHC(pa_region_model1, type = "HC1", cluster = "group")
coef(pa_region_model1,vcov=clustered_se4)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##    -0.004683571     0.047761386    -0.960333546    -0.004144015     0.058879215 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##    -0.037939638     0.788254376     0.497052270     0.066837729
#for pacific south
ps_region<-project1[project1$county=="PS",]
#modelling for pacific south
ps_region_model1<-plm(talremia_inc~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays+tick_lag
                      ,data=ps_region,index=c("county","year"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(ps_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = ps_region, model = "within", index = c("county", 
##     "year"))
## 
## Balanced Panel: n = 1, T = 10, N = 40
## 
## Residuals:
##     Min.  1st Qu.   Median  3rd Qu.     Max. 
## -2.16519 -1.05391 -0.30397  1.02710  3.92342 
## 
## Coefficients:
##                    Estimate  Std. Error t-value Pr(>|t|)  
## CumPrecip        5.9123e-04  3.4276e-03  0.1725  0.86421  
## forest           7.7429e-02  3.0810e-02  2.5132  0.01757 *
## mixeddev        -3.7896e+01  2.0450e+01 -1.8531  0.07372 .
## TempVar         -1.4731e-02  1.3943e-02 -1.0566  0.29914  
## PrecipVar       -5.3794e-02  3.4936e-02 -1.5398  0.13410  
## AvgWinterTemp    2.3126e-02  7.6478e-02  0.3024  0.76444  
## AvgSpringPrecip -6.8617e-01  4.9118e-01 -1.3970  0.17267  
## HotDryDays       1.4974e-01  9.1565e-02  1.6353  0.11244  
## tick_lag        -2.2004e-02  2.9244e-02 -0.7524  0.45766  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    146.4
## Residual Sum of Squares: 85.756
## R-Squared:      0.41424
## Adj. R-Squared: 0.23851
## F-statistic: 2.35724 on 9 and 30 DF, p-value: 0.037781
#cluster robust se
library(sandwich)
clustered_se5 <- vcovHC(ps_region_model1, type = "HC1", cluster = "group")
coef(ps_region_model1,vcov=clustered_se5)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##    5.912342e-04    7.742941e-02   -3.789604e+01   -1.473131e-02   -5.379431e-02 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##    2.312614e-02   -6.861673e-01    1.497359e-01   -2.200407e-02
#for southeast region
SE_region<-project1[project1$county=="SE",]
#modelling for south east region
SE_region_model1<-plm(talremia_inc~CumPrecip+forest
                      +mixeddev
                      +TempVar
                      +PrecipVar
                      +AvgWinterTemp
                      +AvgSpringPrecip
                      +HotDryDays+tick_lag
                      ,data=SE_region,index=c("county","year"),model="within")
## Warning in pdata.frame(data, index = index, ...): duplicate couples (id-time) in resulting pdata.frame
##  to find out which, use, e.g., table(index(your_pdataframe), useNA = "ifany")
summary(SE_region_model1)
## Oneway (individual) effect Within Model
## 
## Call:
## plm(formula = talremia_inc ~ CumPrecip + forest + mixeddev + 
##     TempVar + PrecipVar + AvgWinterTemp + AvgSpringPrecip + HotDryDays + 
##     tick_lag, data = SE_region, model = "within", index = c("county", 
##     "year"))
## 
## Balanced Panel: n = 1, T = 10, N = 120
## 
## Residuals:
##      Min.   1st Qu.    Median   3rd Qu.      Max. 
## -20.80932  -4.20306  -0.55705   2.63197  74.28454 
## 
## Coefficients:
##                   Estimate Std. Error t-value  Pr(>|t|)    
## CumPrecip       -0.0022348  0.0094197 -0.2373    0.8129    
## forest          -0.2752164  0.0573016 -4.8029 4.971e-06 ***
## mixeddev         0.3240405  0.5238704  0.6186    0.5375    
## TempVar          0.0554697  0.0383021  1.4482    0.1504    
## PrecipVar       -0.0708119  0.1177459 -0.6014    0.5488    
## AvgWinterTemp    0.0020237  0.3230981  0.0063    0.9950    
## AvgSpringPrecip  1.0073852  0.8419796  1.1964    0.2341    
## HotDryDays       0.0655043  0.1967337  0.3330    0.7398    
## tick_lag         0.0693735  0.1104660  0.6280    0.5313    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Total Sum of Squares:    18682
## Residual Sum of Squares: 12813
## R-Squared:      0.31414
## Adj. R-Squared: 0.25802
## F-statistic: 5.59807 on 9 and 110 DF, p-value: 2.399e-06
#cluster robust se
library(sandwich)
clustered_se6 <- vcovHC(SE_region_model1, type = "HC1", cluster = "group")
coef(SE_region_model1,vcov=clustered_se6)
##       CumPrecip          forest        mixeddev         TempVar       PrecipVar 
##    -0.002234845    -0.275216448     0.324040509     0.055469680    -0.070811905 
##   AvgWinterTemp AvgSpringPrecip      HotDryDays        tick_lag 
##     0.002023730     1.007385214     0.065504280     0.069373538