4.2 MXHI Files

  READMX is also used to extract data from NEWSIPS high dispersion MXHI files, although the calling sequence is slightly different. The procedure will return the FITS header, a wavelength array, absolutely calibrated fluxes, and (optionally) quality flags, noise vector, background, net, and ripple-corrected net fluxes. When multiple orders are extracted, READMX splices them together at the wavelength (within the overlapping wavelength region) where the difference in the ripple correction of adjacent orders is a minimum. The calling sequence is as follows:

READMX,'IMAGET',H,W,F,q,noise,b,net,r,wrange=wrange, $
orange=orange,/uncalib,/noisecal 

IMAGET	FITS file name (including extension).
H	FITS main header, a string array. The VICAR label will be included.
W	Wavelengths (Å, computed from the starting wavelength and increment stored in the MXLO file.
F	Fluxes (erg cm-2 s-1 Å-1 ).
q	Quality flags (``$\nu$'' flags). These are bit encoded. See Section 7.2.2 for a description of the values.
noise	Noise vector. These are error estimates based on the high dispersion noise model for the pixels in the extraction slit. They are specified in FN units but may be calibrated using the /noisecal keyword described below.
b	Chebyshev fit to the background, scaled to the net fluxes.
net	Net fluxes (``flux numbers'').
r	Ripple-corrected net fluxes (``flux numbers'').
wrange	A 2-element vector giving the starting and ending wavelength (in Angstroms) to be extracted.
orange	A 1- or 2-element vector specifying the range of orders to be extracted. If both orange and wrange are specified, READMX extracts the orders specified by orange, and trims the wavelengths according to wrange. Generally, the keywords should not be specified together. To see the complete spectrum for a given order, use READMX with orange equal to a single order number.
uncalib	By default, READMX removes points which are uncalibrated (i.e., points are removed where the nu flag value is -2). Specifying /uncalib causes all extracted points to be extracted (although, when multiple orders are requested, the orders will still be spliced together
noisecal	By default, the noise vector is returned in FN units (i.e., like the net and ripple-corrected net flux vectors). Specifying /noisecal causes the noise vector for each extracted order to be multiplied by the ratio of the absolute flux to the net flux, thereby giving it the same units as the absolute flux vector. Note: this vector is left uncalibrated because the project felt it was not an accurate estimate of the error in the extracted fluxes. Users are cautioned not to rely to heavily on the calibrated noise vector as an error estimate.

See high dispersion examples of READMX in Table 4.2. Note WRITEMX has not yet been modified to support high dispersion NEWSIPS files.

 

 

Table 4.2: MXHI I/O Examples

Expression	Definition
readmx,'swp20246.mxhi',h,w,f	extract wavelengths and fluxes and be prompted for wavelength range.
readmx,'lwp11367.mxhi',h,w,f,q,noise, $ wrange=[2790,2810]	Read a region of a high dispersion spectrum containing Mg II.
readmx,'lwp12345.mxhi',h,w,f,q,cn, $ orange=100,/uncalib,/noisecal plot,w,f errbar,w,f,w*0,cn,psy=1	Reads all data from order 100, calibrates the noise vector, and plots results with error bars based on the noise vector.