I have not encountered this. But, I have integrated netcdf reading into
some of our applications here and I am somewhat familiar with it.
If you look at ptraj.c and search for "spatialVID" you'll see where this
variable is read and written, and you'll find your error message.
If the 'x', 'y', 'z' is not being read that _IS_ important, in that it
suggests possible file corruption, or data misreading, in some fundamental
way.
Here are more technical notes on how the file is supposed to be laid out.
http://amber.scripps.edu/netcdf/nctraj.html
Ptraj will complain on reading a netcdf file without box information, but
that is a warning, not "a hard bug"
In light of the code and specs, you might want to dump out the start of
netcdf file and see if anything strikes you as immediately odd. Please do
keep us all posted. If there is a bug in ptraj/netcdf, even if specific
to your platform, that is very important!
Here is one example of one of my files:
mybox 124> xxd 151-200.crd | head -50
0000000: 4344 4602 0000 0032 0000 000a 0000 0006 CDF....2........
0000010: 0000 0005 6672 616d 6500 0000 0000 0000 ....frame.......
0000020: 0000 0007 7370 6174 6961 6c00 0000 0003 ....spatial.....
0000030: 0000 0004 6174 6f6d 0000 2297 0000 0005 ....atom..".....
0000040: 6c61 6265 6c00 0000 0000 0005 0000 000c label...........
0000050: 6365 6c6c 5f73 7061 7469 616c 0000 0003 cell_spatial....
0000060: 0000 000c 6365 6c6c 5f61 6e67 756c 6172 ....cell_angular
0000070: 0000 0003 0000 000c 0000 0006 0000 0005 ................
0000080: 7469 746c 6500 0000 0000 0002 0000 0000 title...........
0000090: 0000 000b 6170 706c 6963 6174 696f 6e00 ....application.
00000a0: 0000 0002 0000 0005 414d 4245 5200 0000 ........AMBER...
00000b0: 0000 0007 7072 6f67 7261 6d00 0000 0002 ....program.....
00000c0: 0000 0006 7361 6e64 6572 0000 0000 000e ....sander......
00000d0: 7072 6f67 7261 6d56 6572 7369 6f6e 0000 programVersion..
00000e0: 0000 0002 0000 0003 392e 3000 0000 000b ........9.0.....
00000f0: 436f 6e76 656e 7469 6f6e 7300 0000 0002 Conventions.....
0000100: 0000 0005 414d 4245 5200 0000 0000 0011 ....AMBER.......
0000110: 436f 6e76 656e 7469 6f6e 5665 7273 696f ConventionVersio
0000120: 6e00 0000 0000 0002 0000 0003 312e 3000 n...........1.0.
0000130: 0000 000b 0000 0005 0000 0007 7370 6174 ............spat
0000140: 6961 6c00 0000 0001 0000 0001 0000 0000 ial.............
0000150: 0000 0000 0000 0002 0000 0004 0000 0000 ................
0000160: 0000 0264 0000 0004 7469 6d65 0000 0001 ...d....time....
0000170: 0000 0000 0000 000c 0000 0001 0000 0005 ................
0000180: 756e 6974 7300 0000 0000 0002 0000 000a units...........
0000190: 7069 636f 7365 636f 6e64 0000 0000 0005 picosecond......
00001a0: 0000 0004 0000 0000 0000 027c 0000 000b ...........|....
00001b0: 636f 6f72 6469 6e61 7465 7300 0000 0003 coordinates.....
00001c0: 0000 0000 0000 0002 0000 0001 0000 000c ................
00001d0: 0000 0001 0000 0005 756e 6974 7300 0000 ........units...
00001e0: 0000 0002 0000 0008 616e 6773 7472 6f6d ........angstrom
00001f0: 0000 0005 0001 9f14 0000 0000 0000 0280 ................
0000200: 0000 000c 6365 6c6c 5f73 7061 7469 616c ....cell_spatial
0000210: 0000 0001 0000 0004 0000 0000 0000 0000 ................
0000220: 0000 0002 0000 0004 0000 0000 0000 0268 ...............h
0000230: 0000 000c 6365 6c6c 5f61 6e67 756c 6172 ....cell_angular
0000240: 0000 0002 0000 0005 0000 0003 0000 0000 ................
0000250: 0000 0000 0000 0002 0000 0010 0000 0000 ................
0000260: 0000 026c 7879 7a00 6162 6300 616c 7068 ...lxyz.abc.alph
0000270: 6162 6574 6120 6761 6d6d 6100 0000 0000 abeta gamma.....
0000280: 4287 bdf4 42af e7f0 424e 7efa 4288 4831 B...B...BN~.B.H1
0000290: 42af 8a3d 424a ac08 4287 0419 42b1 c831 B..=BJ..B...B..1
> Dear Amber users,
>
> I have recently been using the NETCDF trajectory format in AMBER and
> decided not only are they easier to handle but process much faster in
> other programmes such as ptraj.
> However, after converting an ASCII trajectory into a NETCDF
> trajectory I find that whenever it is subsequently processed in ptraj
> I get the following message :
>
>
> Whoa Nellie & Cripes!!! We should see 'x', 'y' and 'y' characters in
> the spatial variables of the AMBER NetCDF trajectory file unless
> someone altered the spec. We see '', '' and ''
>
> The processed trajectory does not seem to have problems but I was
> wondering if anyone else has come across this?
>
>
> Dr Geoffrey Wood
> Ecole Polytechnique Fédérale de Lausanne
> SB - ISIC - LCBC
> BCH 4108
> CH - 1015 Lausanne
>
>
>
>
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Aug 26 2007 - 06:07:39 PDT
