Problem with scatter2gridlaplace_xy on Linux

[Lev Tarasoff <lev@xxxxxxxxxxxxxxxxxxxxxxxxxx>]

I'm trying to grid some lat,long, elevation triplets onto a regular grid
and scatter2gridlaplace_xy is giving me mixed up results but with no error
message (ie it appears that most of the points are ignored and values are
wrong). The new objective_analysis_demo works on my machine, but the
following script doesn't even though it matches the form of the demo and
of examples from the mail archives. What's wrong?

DEFINE AXIS/X=74.5W:4.5W:1/UNIT=DEGREE xlongr
DEFINE AXIS/Y=56.75N:84.75N:0.5/UNIT=DEGREE ylatgr

DEFINE AXIS/X=1:83:1 xin
DEFINE AXIS/Y=1:141:1 yin
DEFINE GRID/X=xin/Y=yin  gin

FILE/SKIP=2/VAR="ia,jb,lat,longw" /GRID=gin grid20-EISMINT
FILE/SKIP=2/VAR="elev,H,hp,pre" /GRID=gin suaq20-EISMINT

let xln=360. + longw[d=grid20-EISMINT] ! ensure longitude is > 0.
let ylt= lat[d=grid20-EISMINT]
LET hel = scatter2gridlaplace_xy(xln,ylt,elev,x[gx=xlongr],y[gy=ylatgr],5.,5)

The 2 input files have the form:
grid20-EISMINT:
 83  141 20
 I    J    Lat       Long
 1    1  58.67642   -55.95658
 2    1  58.71147   -55.63363
 3    1  58.74557   -55.30993
 4    1  58.77872   -54.98550
...
83  141  79.87281    12.88865


suaq20-EISMINT:
 83  141 20
 elev      H   B      precip
-3199      0  -3199   .9300
-3217      0  -3217   .9400
...


Thanks,

Lev


Lev Tarasoff -  Dept of Physics, University of Toronto,
                60 St. George St., Toronto, Ontario, CANADA, M5S 1A7
                Tel (416)-946-3019  Fax (416)-978-8905
                email: lev@atmosp.physics.utoronto.ca