问题
I am trying to mine rules from the events of cable modems. Linked is one file of thousands. When I try and run the cspade algorithm on the merged file of all devices (12 million rows) it spends hours chewing through RAM until it uses all 64 GB I have available. So I attempted to run the algorithm on the linked file for just one device. I see the exact same thing happen. Since this sub sample is only 2190 rows I thought this was strange. Can someone explain why Im not seeing results in a timely matter on this small data set?
https://drive.google.com/file/d/0B6VvhxxLVGccVnhDNmVKUE0yaEk/view?usp=sharing
x <- read_baskets(con = "C:/Documents/cspadeData.txt",info=c('sequenceID','eventID'),sep=' ')
as(x, "data.frame")
s1 <- cspade(x, parameter = list(support = 0.01,maxwin=86400),
control = list(memsize=65536,numpart=1,verbose = TRUE))
summary(s1)
EDIT: I received this error.
mining transactions ...Error in cspade(x, parameter = list(support = 0, maxwin = (15 * 60)), :
system invocation failed
In addition: Warning message:
running command '"C:\Windows\system32\cmd.exe" /c "C:/Users/baferrier/Documents/R/win-library/3.2/arulesSequences/bin/x64/spade" -i C:\Users\BAFERR~1\AppData\Local\Temp\2\Rtmp0uwNR8\cspade201c4d2b1cf1 -s 0 -m 65536 -Z 10 -z 10 -w 900 -r -e 1 -o > C:\Users\BAFERR~1\AppData\Local\Temp\2\Rtmp0uwNR8\cspade201c4d2b1cf1.out'
had status 65535
来源：https://stackoverflow.com/questions/37492087/cspade-r-error