R devel - Feb 2008 - Got *** caught segfault *** with Quantreg on Mac (PR#10699)

Full_Name: Edward Huang
Version: 2.6.1
OS: Mac OS 10.5.1 Leopard
Submission from: (NULL) (71.198.106.232)


I'm trying to run quantile regression on my data. I just couldn't make
it work.
The same dataset ran okay on STATA 10, tho.

Would you please take a look at it?

Here is the error message:



 *** caught segfault ***
address 0x3ff00008, cause 'memory not mapped'

Traceback:
 1: .Fortran("rqbr", as.integer(n), as.integer(p), as.integer(n +    
5),
as.integer(p + 3), as.integer(p + 4), as.double(x), as.double(y),    
as.double(tau), as.double(tol), flag = as.integer(1), coef = double(p),    
resid = double(n), integer(n), double((n + 5) * (p + 4)),     double(n),
as.integer(nsol), as.integer(ndsol), sol = double((p +         3) * nsol), dsol
= double(n * ndsol), lsol = as.integer(0),     h = integer(p * nsol), qn
as.double(qn), cutoff = as.double(cutoff),     ci = double(4 * p), tnmat
double(4 * p), as.double(big),     as.logical(lci1), PACKAGE =
"quantreg")
 2: rq.fit.br(x, y, tau = tau, ...)
 3: switch(method, fn = rq.fit.fn(x, y, tau = tau, ...), fnb = rq.fit.fnb(x,    
y, tau = tau, ...), fnc = rq.fit.fnc(x, y, tau = tau, ...),     pfn
rq.fit.pfn(x, y, tau = tau, ...), br = rq.fit.br(x,         y, tau = tau, ...),
fcen = rq.fit.fcen(x, y, tau = tau,         ...), lasso = rq.fit.lasso(x, y, tau
= tau, ...), scad = rq.fit.scad(x,         y, tau = tau, ...), {        what
<-
paste("rq.fit.", method, sep = "")        if (exists(what,
mode = "function"))
          (get(what, mode = "function"))(x, y, ...)        else
stop(paste("unimplemented method:", method))    })
 4: rq.fit(X, Y, tau = tau, method, ...)

 5: rq(PAX ~ PFC + AA + UA + DL + CO + NW + US + Q931 + Q932 + Q933 +     Q934 +
Q941 + Q942 + Q943 + Q944 + Q951 + Q952 + Q953 + Q954 +     Q961 + Q962 + Q963 +
Q964 + ATLBOS + ATLDEN + ATLDFW + ATLDTW +     ATLEWR + ATLIAH + ATLJFK + ATLLAS
+ ATLLAX + ATLLGA + ATLMCO +     ATLMIA + ATLMSP + ATLORD + ATLPHX + ATLPIT +
ATLSEA + ATLSFO +     ATLSTL + BOSDEN + BOSDFW + BOSDTW + BOSEWR + BOSIAH +
BOSJFK +     BOSLAS + BOSLAX + BOSLGA + BOSMCO + BOSMIA + BOSMSP + BOSORD +    
BOSPHX + BOSPIT + BOSSEA + BOSSFO + BOSSTL + DENDFW + DENDTW +     DENEWR +
DENIAH + DENJFK + DENLAS + DENLAX + DENLGA + DENMCO +     DENMIA + DENMSP +
DENORD + DENPHX + DENPIT + DENSEA + DENSFO +     DENSTL + DFWDTW + DFWEWR +
DFWIAH + DFWJFK + DFWLAS + DFWLAX +     DFWLGA + DFWMCO + DFWMIA + DFWMSP +
DFWORD + DFWPHX + DFWPIT +     DFWSEA + DFWSFO + DFWSTL + DTWEWR + DTWIAH +
DTWJFK + DTWLAS +     DTWLAX + DTWLGA + DTWMCO + DTWMIA + DTWMSP + DTWORD +
DTWPHX +     DTWPIT + DTWSEA + DTWSFO + DTWSTL + EWRIAH + EWRJFK + EWRLAS +    
EWRLAX + EWRLGA + EWRMCO + EWRMIA + EWRMSP + EWRORD + EWRPHX +     EWRPIT +
EWRSEA + EWRSFO + EWRSTL + IAHJFK + IAHLAS + IAHLAX +     IAHLGA + IAHMCO +
IAHMIA + IAHMSP + IAHORD + IAHPHX + IAHPIT +     IAHSEA + IAHSFO + IAHSTL +
JFKLAS + JFKLAX + JFKLGA + JFKMCO +     JFKMIA + JFKMSP + JFKORD + JFKPHX +
JFKPIT + JFKSEA + JFKSFO +     JFKSTL + LASLAX + LASLGA + LASMCO + LASMIA +
LASMSP + LASORD +     LASPHX + LASPIT + LASSEA + LASSFO + LASSTL + LAXLGA +
LAXMCO +     LAXMIA + LAXMSP + LAXORD + LAXPHX + LAXPIT + LAXSEA + LAXSFO +    
LAXSTL + LGAMCO + LGAMIA + LGAMSP + LGAORD + LGAPHX + LGAPIT +     LGASEA +
LGASFO + LGASTL + MCOMIA + MCOMSP + MCOORD + MCOPHX +     MCOPIT + MCOSEA +
MCOSFO + MCOSTL + MIAMSP + MIAORD + MIAPHX +     MIAPIT + MIASEA + MIASFO +
MIASTL + MSPORD + MSPPHX + MSPPIT +     MSPSEA + MSPSFO + MSPSTL + ORDPHX +
ORDPIT + ORDSEA + ORDSFO +     ORDSTL + PHXPIT + PHXSEA + PHXSFO + PHXSTL +
PITSEA + PITSFO +     PITSTL + SEASFO + SEASTL + SFOSTL, tau = 0.5)

Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace

Edward,

On Feb 5, 2008, at 4:25 AM, huangce at gmail.com wrote:
> Full_Name: Edward Huang
> Version: 2.6.1
> OS: Mac OS 10.5.1 Leopard
> Submission from: (NULL) (71.198.106.232)
>
>
> I'm trying to run quantile regression on my data. I just couldn't  
> make it work.
> The same dataset ran okay on STATA 10, tho.
>
> Would you please take a look at it?
>
sure, gladly, if you send us the data (or any other way to reproduce  
it) ...

BTW: this is a crash in a contributed package, so you should ask the  
maintainer of the package and not file an R bug report.

Thanks,
Simon

> Here is the error message:
>
>
>
> *** caught segfault ***
> address 0x3ff00008, cause 'memory not mapped'
>
> Traceback:
> 1: .Fortran("rqbr", as.integer(n), as.integer(p), as.integer(n +
> 5),
> as.integer(p + 3), as.integer(p + 4), as.double(x), as.double(y),
> as.double(tau), as.double(tol), flag = as.integer(1), coef =  
> double(p),
> resid = double(n), integer(n), double((n + 5) * (p + 4)),      
> double(n),
> as.integer(nsol), as.integer(ndsol), sol = double((p +         3) *  
> nsol), dsol
> = double(n * ndsol), lsol = as.integer(0),     h = integer(p *  
> nsol), qn > as.double(qn), cutoff = as.double(cutoff),     ci = double(4
* p),
> tnmat > double(4 * p), as.double(big),     as.logical(lci1), PACKAGE =  
> "quantreg")
> 2: rq.fit.br(x, y, tau = tau, ...)
> 3: switch(method, fn = rq.fit.fn(x, y, tau = tau, ...), fnb =  
> rq.fit.fnb(x,
> y, tau = tau, ...), fnc = rq.fit.fnc(x, y, tau = tau, ...),     pfn >
rq.fit.pfn(x, y, tau = tau, ...), br = rq.fit.br(x,         y, tau =
> tau, ...),
> fcen = rq.fit.fcen(x, y, tau = tau,         ...), lasso =  
> rq.fit.lasso(x, y, tau
> = tau, ...), scad = rq.fit.scad(x,         y, tau = tau, ...),  
> {        what <-
> paste("rq.fit.", method, sep = "")        if
(exists(what, mode =
> "function"))
>          (get(what, mode = "function"))(x, y, ...)        else
> stop(paste("unimplemented method:", method))    })
> 4: rq.fit(X, Y, tau = tau, method, ...)
>
> 5: rq(PAX ~ PFC + AA + UA + DL + CO + NW + US + Q931 + Q932 + Q933  
> +     Q934 +
> Q941 + Q942 + Q943 + Q944 + Q951 + Q952 + Q953 + Q954 +     Q961 +  
> Q962 + Q963 +
> Q964 + ATLBOS + ATLDEN + ATLDFW + ATLDTW +     ATLEWR + ATLIAH +  
> ATLJFK + ATLLAS
> + ATLLAX + ATLLGA + ATLMCO +     ATLMIA + ATLMSP + ATLORD + ATLPHX +  
> ATLPIT +
> ATLSEA + ATLSFO +     ATLSTL + BOSDEN + BOSDFW + BOSDTW + BOSEWR +  
> BOSIAH +
> BOSJFK +     BOSLAS + BOSLAX + BOSLGA + BOSMCO + BOSMIA + BOSMSP +  
> BOSORD +
> BOSPHX + BOSPIT + BOSSEA + BOSSFO + BOSSTL + DENDFW + DENDTW +      
> DENEWR +
> DENIAH + DENJFK + DENLAS + DENLAX + DENLGA + DENMCO +     DENMIA +  
> DENMSP +
> DENORD + DENPHX + DENPIT + DENSEA + DENSFO +     DENSTL + DFWDTW +  
> DFWEWR +
> DFWIAH + DFWJFK + DFWLAS + DFWLAX +     DFWLGA + DFWMCO + DFWMIA +  
> DFWMSP +
> DFWORD + DFWPHX + DFWPIT +     DFWSEA + DFWSFO + DFWSTL + DTWEWR +  
> DTWIAH +
> DTWJFK + DTWLAS +     DTWLAX + DTWLGA + DTWMCO + DTWMIA + DTWMSP +  
> DTWORD +
> DTWPHX +     DTWPIT + DTWSEA + DTWSFO + DTWSTL + EWRIAH + EWRJFK +  
> EWRLAS +
> EWRLAX + EWRLGA + EWRMCO + EWRMIA + EWRMSP + EWRORD + EWRPHX +      
> EWRPIT +
> EWRSEA + EWRSFO + EWRSTL + IAHJFK + IAHLAS + IAHLAX +     IAHLGA +  
> IAHMCO +
> IAHMIA + IAHMSP + IAHORD + IAHPHX + IAHPIT +     IAHSEA + IAHSFO +  
> IAHSTL +
> JFKLAS + JFKLAX + JFKLGA + JFKMCO +     JFKMIA + JFKMSP + JFKORD +  
> JFKPHX +
> JFKPIT + JFKSEA + JFKSFO +     JFKSTL + LASLAX + LASLGA + LASMCO +  
> LASMIA +
> LASMSP + LASORD +     LASPHX + LASPIT + LASSEA + LASSFO + LASSTL +  
> LAXLGA +
> LAXMCO +     LAXMIA + LAXMSP + LAXORD + LAXPHX + LAXPIT + LAXSEA +  
> LAXSFO +
> LAXSTL + LGAMCO + LGAMIA + LGAMSP + LGAORD + LGAPHX + LGAPIT +      
> LGASEA +
> LGASFO + LGASTL + MCOMIA + MCOMSP + MCOORD + MCOPHX +     MCOPIT +  
> MCOSEA +
> MCOSFO + MCOSTL + MIAMSP + MIAORD + MIAPHX +     MIAPIT + MIASEA +  
> MIASFO +
> MIASTL + MSPORD + MSPPHX + MSPPIT +     MSPSEA + MSPSFO + MSPSTL +  
> ORDPHX +
> ORDPIT + ORDSEA + ORDSFO +     ORDSTL + PHXPIT + PHXSEA + PHXSFO +  
> PHXSTL +
> PITSEA + PITSFO +     PITSTL + SEASFO + SEASTL + SFOSTL, tau = 0.5)
>
> Possible actions:
> 1: abort (with core dump, if enabled)
> 2: normal R exit
> 3: exit R without saving workspace
> 4: exit R saving workspace
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>