"Error in value[[3L]](cond) : unable to open database file" occurs frequently

[OhStan]

Hi,

When I was trying to use sqldf to join 2 data frames, it some times returned with error message "Error in value[3L] : unable to open database file", then stops there, nothing else given. Do you know what is going on?

Note:
RSQLite was used.
This can be resolved by commenting out some of the fields from data frames to be joint.
I have converted all columns to "character" data type before calling sqldf

Thanks,

[MaxCntt]

Hi, any news on this one. Since several days, I systematically get the message "Error: unable to open database file" when trying to join two tables. I'm using R memory, no external database.

My code simply look like this:
sqldf("select df1.col1, df2.col2 from df1 left join df2 on df1.col1 = df2.col2")

Thanks

[ggrothendieck]

Have never seen this. Maybe you think you are using sqlite but are actually using something else and it needs a database? Add verbose=TRUE to the sqldf call and see if that shows anything helpful.

[OhStan]

This seems to be kind of random, and I guess it may be related to the SQLite database not initiated properly. If it can return more info then it will certainly help us to figure out the exact cause.

[MaxCntt]

Thanks for the fast reply and the verbose tip.
It seems to use SQLite:

sqldf("select df.Id, df.Name, count(SC.CaseNumber) as Number_of_Case from df left join SC on df.Id = SC.AccountId group by df.Id", verbose=T)

sqldf: library(RSQLite)
sqldf: m <- dbDriver("SQLite")
sqldf: connection <- dbConnect(m, dbname = ":memory:")
sqldf: initExtension(connection)
sqldf: dbWriteTable(connection, 'df', df, row.names = FALSE)
sqldf: dbWriteTable(connection, 'SC', SC, row.names = FALSE)
sqldf: dbGetQuery(connection, 'select df.Id, df.Name, count(SC.CaseNumber) as Number_of_Case
from df left join SC on df.Id = SC.AccountId group by df.Id')
Error: unable to open database file
sqldf: dbDisconnect(connection)

As we can see, both tables are written in the database and I'm able to query them individually. The error occurs with the join.