Test data for for JCAMP-DX software

The following files are made available for people who wish to test their JCAMP-DX programs. Some have come from Bob McDonald, others from Tony Davies, Peter Lampen, Scott Van Bramer and some from our labs here in Jamaica.

The JCAMP-DX standard allows for a variety of file formats which include:

Files containing equally spaced abscissa (X) values.
##XYDATA= (X++(Y..Y))
FIXed format files in increasing or decreasing abscissa order.
PACed format " w or w/out DUP "
SQZ format " w or w/out DUP "
DIF format " w or w/out DUP "
and non-compressed files containing a simple column of XY pairs.

Non-equally spaced X values but continuous plot
##XYPOINTS= (XY..XY)
and tables
##PEAKTABLES= (XY..XY)
among others.

Error free files

Examples of these (supposedly error free) files are given here:
If you do find errors in any of these files please let me know.

fixinc1.jdx (IR-T)
fixdec1.jdx (IR-T)
Fixed format, increasing and decreasing abscissa respectively.

fixinc2.jdx (IR-A)
fixdec2.jdx (NMR)
fixinc3.jdx - a sine curve (1-360).
fixdec3.jdx - a sine curve (360-1).
fixinc4.jdx - a simple curve.
fixinc5.jdx - a simulated titration curve.
Fixed format, increasing and decreasing abscissa respectively.

jtpolys.jdx (IR-T)
Fixed format, increasing abscissa.

pacdec1.jdx (IR-T)
PACed format, decreasing abscissa. Note that the difference between FIX and PAC is that no multiple blanks are between Y values. In the example given, no blanks at all are between the Y values either a '+' or a '-' is used as the delimiter.

I don't have any examples of PACDUP files, anyone out there seen any??

sqzdec1.jdx (NMR)
An example of a file that uses only the SQZ compression type. Note that this file has a very long header record and a large number of data points so may take a few minutes to load.

sqzdupd1.jdx (IR-T)
An example of a file that contains both SQZ and DUP delimiters.

dupinc1.jdx (IR-A)
dupdec1.jdx (IR-T)
DIFDUP format, increasing and decreasing abscissa respectively.

dupinc2.jdx (IR-T)
dupdec2.jdx (IR-T)
DIFDUP format, increasing and decreasing abscissa respectively.

jtpolysd.jdx (IR-T)
DIFDUP format, increasing abscissa.

xyinc1.jdx (IR-T)
xyinc2.jdx (EPR)
Single column of XY pairs, increasing abscissa.

pktab1.jdx (MS)
pktab2.jdx (MS)
Examples of PEAK TABLES, non continuous display.

Multiple Block Files

CHIME vs 2 will cope with multiple BLOCKS and an extra menu option is available to select which BLOCK is to be displayed.

I include here some files with MULTIPLE BLOCKS:

blckpac1.jdx (Vis)
An example of a kinetic run using PAC'd Vis spectra (5 data blocks).

blckpkt1.jdx (MS)
An example of a series of Mass Spectra recorded with different Ionization energies (6 data blocks).

compound.jdx (IR-T)
An example of some FTIR's in DIFDUP format (5 data blocks).

NTUPLE files

CHIME vs 2 will cope with NTUPLES and an extra menu option is available to select which PAGE is to be displayed.

NMR data, especially from Bruker instruments, have often been recorded using the NTUPLE format defined in the vs 5.0 specs for JCAMP-DX.
The following 14 data files are for o-dichlorobenzene. The first 10 are spectra and show real (1-5) and complex (6-10) spectra at different levels of compression. The remaining 4 are FID's.

Spectra
o01.jdx  FIX - real only	o06.jdx  FIX - complex (real and imaginary)
o02.jdx  DIF			o07.jdx  DIF
o03.jdx  PAC			o08.jdx  PAC
o04.jdx  SQZ			o09.jdx  SQZ
o05.jdx  DIFDUP			o10.jdx  DIFDUP

FID's
ofid1.jdx  FIX			ofid3.jdx  SQZ
ofid2.jdx  PAC			ofid4.jdx  DIFDUP

Files with errors

Some examples of files that contain common errors such as incorrect number of points given in the header, a line missing, or invalid data are given below.

Note that ##T_ITLE=, ##TITLe=, ##T\I/T_L e= should ALL be interpreted as ##TITLE=. That is, when reading the file you must strip out spaces hyphens, underscores and convert to UPPERCASE before comparison of the labels. ##JCAMP_DX=, ##JCAMP-DX= and ##JCAMPDX= are again, all allowed and equivalent.

efim1.jdx
efip1.jdx
This is file fixinc2.jdx but with 1 less or more point listed in the header than is actually present in the file.

efifx1.jdx - increasing X
efilx1.jdx
efdfx1.jdx - decreasing X
efdlx1.jdx
These files have the header values for FIRSTX and LASTX given incorrectly.

ecp2fe.jdx
There are no JCAMP-DX specifications for cyclic voltammetry but the following gives an example of what might be possible. Since the scan is cyclic from 0 volt back to 0 volt. FIRSTX and LASTX are identical and do NOT match the maximum and minimum values. The only way I could get to display the file was by setting INCORRECT values for FIRSTX and LASTX and reading in the data as ##XYPOINTS.

efibi1.jdx
ttest.jdx
fixed format but contains spaces, $$ comments in data lines and invalid data symbol (?) included.

efidp1.jdx
This has one data point out by a factor of 10000. The resulting graph looks like it has only 1 point, but try a zoom.

dtest.jdx
DIFDUP format but contains spaces, $$ comments in data lines and invalid data symbol (?) included.

Carriage Return, Line Feed Problems

Bob McDonald has summarised the problem of END-OF-LINE (NEW-LINE) DETECTION:

I quote
"Section 4.3 of V4.24 indicates that JCAMP-DX lines should be terminated by < crlf >. In MSDOS, < crlf > is the standard, but UNIX uses < lf > alone, and the Macintosh uses < cr > alone. Actually, most PC software handles either < crlf > or < lf >, but not < cr >. It is not convenient to use non-standard end-of-line on any of these systems. Some data transfer protocols handle the conversion between conventions. Documentation should warn prospective users of this problem so that they can choose the appropriate option of the data transmission protocol or disk conversion program".

mactab1.jdx
mactab2.jdx
These files were copied to a MAC as ASCII text and then sent back as BINARY. This meant that all < crlf >'s were converted to < cr > alone. The plugin should read them successfully, since we stream the data.

Note that the Bruker program, WinNMR, despite the fact that it runs under Windows, does not use the < crlf > convention. The files therefore can NOT be easily edited using a simple ASCII editor like Notepad since they contain < lf > as end of line markers. The argument is that most Bruker users are Unix based so the files are written to conform to Unix.

If you encounter difficulties with software reading Bruker files then try converting from < lf > to < crlf >. One way to do this easily is to FTP them to a UNIX machine and FTP them back again. The sequence must be to FTP them as BINARY then FTP them back as ASCII text.

Created Feb 1997. Last modified 18th November-98.
URL http://wwwchem.uwimona.edu.jm:1104/spectra/testdata/index.html