R version 3.3.2, Windows 10: gstat package: Error in fitting a variogram model using 'fit.variogram' function

Question

I am having some problem with fit.variogram function of the gstat package. My code, data, error message and traceback() message can be found below. Any suggestion to approach this problem would be appreciated.

library(gstat)
mydf<-as.data.frame(read.table("Sample_data.xlsx",header = TRUE,sep = 
",",na.strings = "EMPTY"))
coordinates(mydf) = ~x+y
c.vgm.exp<-variogram(z~1,data=mydf,cutoff=60000, width = 60000/15)
c.vgm.fit<-fit.variogram(c.vgm.exp,vgm(nugget=1000000,psill=6000000,model 
="Exp",range = 40000),fit.method = 7)

Data file can be obtained from: https://gtvault-my.sharepoint.com/personal/srathore6_gatech_edu/_layouts/15/guestaccess.aspx?docid=0b5983d45c9544203a5081cd501c64de4&authkey=AYhC8_ZVvGFG18JHHWMgUa4&expiration=2017-12-01T19%3a45%3a23.000Z

Error Message:

c.vgm.fit<-fit.variogram(c.vgm.exp,vgm(nugget=1000000,psill=6000000,model 
="Exp",range = 40000),fit.method = 7)
Error in switch(model, exponential = fit.exponential(v.object, c0 = nugget,: 
EXPR must be a length 1 vector

traceback()
1: fit.variogram(c.vgm.exp, vgm(nugget = 1e+06, psill = 6e+06, model = 
"Exp", range = 40000), fit.method = 7)