Gridding scattered data

["Newcomb, Donald" <NewcombD@xxxxxxxxxxxxxxxxxxxxx>]

I have an ASCII data file which consists of ~31K more-or-less randomly
located measurments within a 3D volume. The format is similar to:

longitude latitude depth value
90.365 26.112 1239.67 852.9
86.5 24 100 5
....

Where any of the values can have any degree of precision. The data are
sparse and do not fall on regular grid boundaries. I have tried to read the
file into a 3D Ferret grid but the results look like something got out of
phase on the TV screen. I did a "wc" to check the file and the number of
words is 4x the number lines. The Go file I'm using to read the ASCII data
looks like:

DEFINE AXIS/X={limits & step}/UNITS=degrees long
DEFINE AXIS/Y={limits & step}/UNITS=degrees lat
DEFINE AXIS/Z={limits & step}/UNITS=meters deep
DEFINE GRID/X=long/Y=lat/Z=deep datacube
FILE/VAR="long,lat,deep, val"/GRID=datacube "ascii_filename"
EXIT

This is a modification of a Go file I used to read in a large, regularly
spaced, ASCII data file. I  think the question I'm trying to ask is, "Is
there a problem doing this to read irregularly spaced data into a Ferret
grid?"  Is there a good way to  this type data into Ferret? I'm using v4.11.
Thanks.