Random Try
12 June, 2015
Data calling
setwd("/Users/subasishdas1/Desktop/MacPro_File/")
data<-read.csv(file="FFV_counts.csv",head=TRUE,sep=",")
names(data)## [1] "year"
## [2] "article_id"
## [3] "state"
## [4] "Population"
## [5] "Economy"
## [6] "Car.Ownership"
## [7] "Truck.Ownership"
## [8] "Gasoline.Prices"
## [9] "Gas_lag"
## [10] "Legitimation"
## [11] "Entrepreneurial.Experimentation"
## [12] "Resource.Mobilization"
## [13] "Knowledge.development.and.diffusion"
## [14] "Know_lag"
## [15] "Development.of.Positive.External.Economies"
## [16] "Market.Formation"
## [17] "Influence.the.Direction.of.Search"
## [18] "Abbrev"
## [19] "FFV.Count"
## [20] "MY"
## [21] "EthCons"
## [22] "EthCons_bbl"
## [23] "EthProd"
## [24] "FFVStations"
## [25] "CumStations"
## [26] "RFS"
## [27] "AFV.Manufacturer.Retrofitter"
## [28] "Acquisition...Fuel.Use"
## [29] "Aftermarket.Conversions"
## [30] "Air.Quality...Emissions"
## [31] "Alternative.Fuel.Dealer"
## [32] "Alternative.Fuel.Producer"
## [33] "Alternative.Fuel.Purchaser"
## [34] "Climate.Change...Energy.Initiatives"
## [35] "Driving...Idling"
## [36] "Ethanol"
## [37] "Exemptions"
## [38] "Fleet.Purchaser.Manager"
## [39] "Fuel.Economy...Efficiency"
## [40] "Fuel.Production...Quality"
## [41] "Fuel.Taxes"
## [42] "Fueling...TSE.Infrastructure.Owner"
## [43] "Grants"
## [44] "Idle.Reduction"
## [45] "Loans.and.Leases"
## [46] "Rebates"
## [47] "Registration...Licensing"
## [48] "Renewable.Fuel.Standard...Mandate"
## [49] "Tax.Incentives"
## [50] "Vehicle.Owner.Driver"
## [51] "Time"
## [52] "Time2"Count data modeling
F1<-glm(Legitimation~AFV.Manufacturer.Retrofitter+
Acquisition...Fuel.Use+
Aftermarket.Conversions+
Air.Quality...Emissions+
Alternative.Fuel.Dealer+
Alternative.Fuel.Producer+
Alternative.Fuel.Purchaser+
Climate.Change...Energy.Initiatives+
Exemptions+Fleet.Purchaser.Manager+
Fuel.Economy...Efficiency+
Fuel.Production...Quality+
Fuel.Taxes+
Fueling...TSE.Infrastructure.Owner+
Grants+Idle.Reduction+Loans.and.Leases+
Rebates+Registration...Licensing+
Renewable.Fuel.Standard...Mandate+
Tax.Incentives+Vehicle.Owner.Driver+
Car.Ownership,
data=data,family=quasipoisson)
summary(F1)##
## Call:
## glm(formula = Legitimation ~ AFV.Manufacturer.Retrofitter + Acquisition...Fuel.Use +
## Aftermarket.Conversions + Air.Quality...Emissions + Alternative.Fuel.Dealer +
## Alternative.Fuel.Producer + Alternative.Fuel.Purchaser +
## Climate.Change...Energy.Initiatives + Exemptions + Fleet.Purchaser.Manager +
## Fuel.Economy...Efficiency + Fuel.Production...Quality + Fuel.Taxes +
## Fueling...TSE.Infrastructure.Owner + Grants + Idle.Reduction +
## Loans.and.Leases + Rebates + Registration...Licensing + Renewable.Fuel.Standard...Mandate +
## Tax.Incentives + Vehicle.Owner.Driver + Car.Ownership, family = quasipoisson,
## data = data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -51.115 -14.955 -9.810 1.134 128.190
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.9459327 0.0919407 53.795 < 2e-16
## AFV.Manufacturer.Retrofitter -0.1929479 0.1720954 -1.121 0.262536
## Acquisition...Fuel.Use -0.0732242 0.1893404 -0.387 0.699052
## Aftermarket.Conversions 0.2341197 0.1537874 1.522 0.128294
## Air.Quality...Emissions 0.6258295 0.1962909 3.188 0.001484
## Alternative.Fuel.Dealer 0.2183758 0.1871896 1.167 0.243701
## Alternative.Fuel.Producer -0.0943103 0.1677529 -0.562 0.574131
## Alternative.Fuel.Purchaser -0.2736143 0.1577009 -1.735 0.083103
## Climate.Change...Energy.Initiatives 1.0149581 0.2273686 4.464 9.14e-06
## Exemptions 0.5470088 0.1946693 2.810 0.005070
## Fleet.Purchaser.Manager 0.2441055 0.1919823 1.271 0.203902
## Fuel.Economy...Efficiency 0.2272381 0.1818774 1.249 0.211865
## Fuel.Production...Quality 0.4158005 0.1515572 2.744 0.006207
## Fuel.Taxes 0.5557059 0.1434652 3.873 0.000116
## Fueling...TSE.Infrastructure.Owner -0.1849631 0.1910904 -0.968 0.333354
## Grants 0.1731920 0.1578056 1.098 0.272735
## Idle.Reduction 0.2185462 0.1892679 1.155 0.248544
## Loans.and.Leases 0.3964675 0.2246765 1.765 0.077991
## Rebates -0.4405115 0.2274441 -1.937 0.053105
## Registration...Licensing 0.1446801 0.1525032 0.949 0.343044
## Renewable.Fuel.Standard...Mandate 0.4887929 0.1938588 2.521 0.011873
## Tax.Incentives -0.1870854 0.1781714 -1.050 0.294005
## Vehicle.Owner.Driver 0.5461028 0.1612267 3.387 0.000739
## Car.Ownership 0.0002742 0.0002000 1.371 0.170798
##
## (Intercept) ***
## AFV.Manufacturer.Retrofitter
## Acquisition...Fuel.Use
## Aftermarket.Conversions
## Air.Quality...Emissions **
## Alternative.Fuel.Dealer
## Alternative.Fuel.Producer
## Alternative.Fuel.Purchaser .
## Climate.Change...Energy.Initiatives ***
## Exemptions **
## Fleet.Purchaser.Manager
## Fuel.Economy...Efficiency
## Fuel.Production...Quality **
## Fuel.Taxes ***
## Fueling...TSE.Infrastructure.Owner
## Grants
## Idle.Reduction
## Loans.and.Leases .
## Rebates .
## Registration...Licensing
## Renewable.Fuel.Standard...Mandate *
## Tax.Incentives
## Vehicle.Owner.Driver ***
## Car.Ownership
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 681.9619)
##
## Null deviance: 518436 on 866 degrees of freedom
## Residual deviance: 326468 on 843 degrees of freedom
## (51 observations deleted due to missingness)
## AIC: NA
##
## Number of Fisher Scoring iterations: 6plot(F1)
Correlation plot
#setwd("/Users/subasishdas1/Desktop/MacPro_File/")
#data<-read.csv(file="FFV_counts.csv",head=TRUE,sep=",")
data3 <- data[c(10, 27: 34, 37: 50, 6)]
names(data3)## [1] "Legitimation"
## [2] "AFV.Manufacturer.Retrofitter"
## [3] "Acquisition...Fuel.Use"
## [4] "Aftermarket.Conversions"
## [5] "Air.Quality...Emissions"
## [6] "Alternative.Fuel.Dealer"
## [7] "Alternative.Fuel.Producer"
## [8] "Alternative.Fuel.Purchaser"
## [9] "Climate.Change...Energy.Initiatives"
## [10] "Exemptions"
## [11] "Fleet.Purchaser.Manager"
## [12] "Fuel.Economy...Efficiency"
## [13] "Fuel.Production...Quality"
## [14] "Fuel.Taxes"
## [15] "Fueling...TSE.Infrastructure.Owner"
## [16] "Grants"
## [17] "Idle.Reduction"
## [18] "Loans.and.Leases"
## [19] "Rebates"
## [20] "Registration...Licensing"
## [21] "Renewable.Fuel.Standard...Mandate"
## [22] "Tax.Incentives"
## [23] "Vehicle.Owner.Driver"
## [24] "Car.Ownership"dim(data3)## [1] 918 24### Using complete cases without NA
data4 <- data3[complete.cases(data3),]
dim(data4)## [1] 867 24### checking correlation
library(corrplot)
corrp <- cor(data4)
corrplot(corrp, method = "circle")
Important variable
### Using randomforest to find important variables
library(randomForest)## randomForest 4.6-10
## Type rfNews() to see new features/changes/bug fixes.data4.rf <- randomForest(Legitimation ~ ., data=data4, ntree=100, keep.forest=FALSE, importance=TRUE)
varImpPlot(data4.rf)
Algorithmic Modeling
index <- 1:nrow(data4)
testindex <- sample(index, trunc(length(index)/3))
testset <- data4[testindex,]
trainset <- data4[-testindex,]
dim(testset)## [1] 289 24dim(trainset)## [1] 578 24library(e1071)
svm.model <- svm(Legitimation ~ ., data = data4, cost = 100, gamma = 1)
svm.pred <- predict(svm.model, testset[,-1])
### Original vs. Predicted plot
plot(testset$Legitimation, svm.pred)