%% datafile        1    4216.38    4219.11
 %% datafile        1    4224.34    4225.28
 %% datafile        1    3307.91    3313.32
  C IV    1.724340aa 0.000006   16.37     1.59  12.522    0.061  0    3     1.2141396  182  156  0.035  0
  C IV    1.724465   0.000008    3.59     2.48  11.745    0.168  0
  C IV    1.724652   0.000075   27.82    15.08  12.158    0.267  0
  H I     1.724340AA 0.000000   31.53     1.46  14.827    0.056  0
  H I     1.724066   0.000039   57.72     1.84  14.151    0.062  0
  H I     1.722318   0.000019   35.59     3.59  12.844    0.047  0
  <>      1.723284SZ 0.000000  -13.94     9.23   1.010    0.007  0
  H I     1.722943   0.000008    4.77     1.66  12.118    0.076  0
  <>      2.469482SZ 0.000000    5.54    11.39   1.002    0.003  0
 
 

The first three lines give the filenames, spectrum number (needed for multispectrum files), start and end wavelengths for the regions fitted. The '%%' is just to tell the program that it is a filename line if you choose to restart VPFIT from one of these files. The fitted parameters are shown for each ion in each system on the subsequent lines. The quantities are, in order

(1) element and ionization level, separated by a space if the element is denoted by a single character.
(2) & (3) redshift and one-sigma error estimate
(4) & (5) Doppler parameter (b = sqrt(2) times Gaussian sigma) and one-sigma error estimate
(6) & (7) log ion column density (cm^{-2}) and one-sigma error estimate (these may be linear values if specified in the program when run)
(8)  Almost invariably zero. if nonzero, apply the ion on this line only to the region whose number (in the order above) is the number given here. If this number is zero, apply all lines from the ion in all relevant given regions. This is rarely used, but allows continuum correction (for example) where the same line appears in two different spectra of the same wavelength region and the spectra are not combined (e.g. because the resolution is different, or you are worried about the noise characteristics).

The first ion has further information concerning the fit:

(9) The number of iterations required to obtain a satisfactory fit.
(10) The normalized chi^2 for the overall fit.
(11) The total number of pixels in the regions fitted
(12) The number of degrees of freedom for the fit.
(13) The approximate probability that the fit given properly represents the data (from Chi^2)
(14)  Is number of dropped systems cf previous iteration (I can't remember why that is there)
 

The parameters for the continuum adjustment '<>' line are similar in style to those for ions:
(2) & (3) redshift, z, (which must be fixed) and one-sigma error estimate (which is consequently zero). This is used to set a reference wavelength for the continuum parameters, which is (1+z) times the wavelength assigned to the <> entry in the atom.dat file. (normally 1215.67, the same as Ly-a).
(4) & (5) continuum slope (<linterm>) and estimated error
(6) & (7) continuum level (<level>) at the reference wavelength and error estimate.

The continuum within the region which contains the wavelength 1215.67(1+z) is then multiplied by  (<level> + <linterm>* (wavelength/reference wavelength - 1), so for small adjustments <level> is somewhere vaguely near unity and <linterm> tends to be in single or low double figures.
 

It the Doppler parameters are tied the output looks a little different, since the turbulent component (zero if free) and /or temperature (again zero if free) have to be specified. So you might see something like:

 %%  filename        1    6399.06    6401.54
 %%  filename        1    5847.67    5849.28
  AlII    2.830577A  0.000000    5.05     0.31  12.177    0.021  0    4     1.0216554  102   81  0.425  0
  AlII    2.830406B  0.000000    2.30BA   0.00  11.621    0.041  0
  SiII    2.830406b  0.000003    2.30ba   0.36  12.856    0.031  0     0.00 1.000E+00
  SiII    2.830577a  0.000002    3.28     0.36  13.399    0.033  0

Here it was decided that the Doppler parameters for the AlII and SiII at z=2.830406 should be the same, so a low temperature (1K) was given. This appears as a turbulent and temperature against the variable component as shown here. If this file is used to restart the program, the turbulent component will be free to vary, since it is zero in the input file, and the temperature fixed at 1 degree Kelvin. [Zero, zero fixes the turbulent component as zero, and allows the temperature to vary.]