MCTL help

  • MCTL
    • Reverse Polish notation
  • General purpose commands
    • CReate command
    • EXE command
    • RM command
    • PCReate command
    • PRM command
    • FIT command
    • FIx command
    • FList command
    • FRee command
    • FREL command
    • SFIT command
    • VERB command
  • Mössbauer tools
    • MCHECK command
    • MCReate command
    • MRM command
    • REad command
    • FEcal command
    • FOld command
    • INfo command
    • MOIre command
    • VK command
  • Mössbauer fit commands
    • DMATrix command
    • MAtrix command
    • CMATRIX command
    • MFit command
    • MFIX command
    • MDis command
    • MFRee command
    • MList command
    • MRel command
    • MXport command
  • Nuclear resonant scattering fit routine
    • TDIS command
    • TFIT command
    • TFINPUT command
    • TFIX command
    • TFREE command
    • TLIST command
    • TREL command
    • TVAR command

MCTL

Reverse Polish notation

Reverse Polish notation (as used in EXE, FREL, MREL).

Expressions consist of the following elements, separated by spaces:
  1. unsigned numbers
      to form a negative number, use '0 <value> -'
    
  2. symbolic, predefined constants
      cpi= 3.14159265358979323846
      ce= 1.6021892e-19  (elementary charge)
      ck= 1.380662e-23   (Boltzmann's constant)
      cc= 2.99792458e8   (speed of light)
      ch= 6.626176e-34   (Planck's constant)
    
  3. functions
      sin,  cos,  tan,  asin,  acos,  atan,  exp,  ln,  log,  sqrt
      (ln is natural logarithm, log is logarithm base 10)
      (older forms are fsin, fcos, ... )
    
  4. operators
      + - / *
    
  5. parameter references
      'p' followed by <index>, example: p0
    
  6. parameter input from keyboard (not in FREL, MREL)
      'r' followed by <index>, optionally followed by 'q'<question>.
      Example: 'r1 qamplitude' will ask and set p[1] and dp[1]
    
    Example:
      cpi 16 ln / sqrt p1 *
      is evaluated as: p1 * sqrt ( PI / log(16) )

General purpose commands


CReate

    CReate <name> <length>

      Define a 'general purpose' data-space.
      It allocates four CIO arrays of type 'float' and size 'length'.
      You have to supply:
        name   : data-space name (6 characters max)
        length : size of the data-space in channels

      The arrays allocated are:
        name+'d'  (data)
        name+'e'  (errors)
        name+'t'  (fit)
        name+'v'  (x coordinates)
See also RM, PCREATE and PRM.

EXE

    EXE <parameters> <expression>

      Calculates an expression in reverse Polish notation.

      If parameters are used in the expression, error estimates of the result
      are calculated, based on the covariance matrix (cov) if available and/or
      the error estimates (dp) of the parameters.
      You have to supply:
        parameters : parameter space to use.
                     type <return> or ',,' if no parameters will be used
        expression : the expression, terminated by end-of-line or ';'
                     parameters p0, p1 ... may be used in the expression

      Example:
        exe sp p1 p9 * cpi *
        will calculate: p1 * p9 * 3.1415926536
For details of the syntax, give HELP RPN.

RM

    RM <name>

      Deletes a 'general purpose' data-space.
      The four arrays become undefined.
      You have to supply:
        name : name as given with CREATE
See also CREATE, PCREATE and PRM.

PCReate

    PCReate <name> <nr_par>

      Defines a 'general purpose' parameter-space.
      You have to supply:
        name   : parameter-space name (6 characters max.)
        nr_par : Size of the parameter-space
      It allocates the following CIO arrays, of type 'double':
        name+'p'
        name+'dp'
See also PRM, CREATE and RM.

PRM

    PRM <name>

      Deletes a 'general purpose' parameter-space.
      The two arrays become undefined.
      You have to supply:
        name : parameter-space name, given with PCREATE.
See also PCREATE, CREATE and RM.

FIT

    FIT <d_space> <p_space> <th_nr> <from> <to> [<np>] <init_val> ...

      Fits an array to one of a number of possible functions.
      You have to supply:
        d-space  : the data space, as created by CREATE
                   the data-, error-, and x-arrays of this data space should
                   be filled with meaningful values
        p-space  : the parameters to use in the fit, as created by PCREATE
        th-nr    : an integer number determining the theory
        from,to  : two integers giving the lower and upper channel number of
                   the fitting region in the arrays of the data space
        np       : number of parameters to be used; only asked if needed
                   (in the polynomial this determines the rank)
        init-val : the initial values of the parameters

      A fit that converges too slowly may be interrupted by typing <CTRL-C>
      (press key 'CTRL' and key 'C' simultaneously)

      Theory 1 :  polynomial          y = p0 + p1*x + p2*x^2 ....
      Theory 2 :  cosine              y = p0*cos((x+p2)/p1) + p3
      Theory 3 :  exponential         y = p0 + p1*exp(-x/p2)
      Theory 4 :  double exponential  y = p0 + p1*exp(-x/p2) + p3*exp(-x/p4)
      Theory 5 :  trapezium function  [ ??? not functional, debugging required ]
        fits position and width of channeling dips
        p0=shoulder height (in counts)
        p1=depth (relative to p0)
        p2=centre (in "channel numbers")
        p3=width at half depth (in "channel numbers")
        p4=width of the left slope (relative to total width)
        p5=width of the right slope(relative to total width)
      Theory 6 :  A = f(t_a) for Mössbauer spectra
        p0=GAMMA_a*f_s/2 (either in mm/s or eV's)
        p1=Debye temperature (in K)
        p2=sigma_0 cross-section (in cm^-2)
        p3=GAMMA_N Heisenberg linewidth (same units as GAMMA_a!!!)
        p4=E_r/k_B recoil energy/k_Boltzman
           (E_e in eV's and k_B = 8.617 10^-5 eV/K)
        p5=N_a number Mössbauer atoms in the absorber (atoms/cm^2)
        p6=GAMMA_a linewidth of the absorption spectrum 
           (the same units as GAMMA_N!!!)
        NOTE. Keep p2-6 fixed.

FIx

    FIx <i> <j> <k> ...

      Prevents the parameters p[i],p[j],p[k],... in the 'general' fit routine,
      from being changed in the fit-routine.
See also FREE, FREL and FLIST.

FList

    FList

      Lists all parameters, in the 'general' fit routine, which are fixed
      and/or have a relation defined. 
See also FREE, FIX and FREL.

FRee

    FRee <i> <j> <k> ...;

      Frees the parameters p[i],p[j],p[k],... in the 'general' fit routine,
      and releases constraints set by FIX or FREL.
See also FIX, FREL and FLIST.

FREL

    FREL <i> <relation for p[i]>

      Defines a relation between the parameters used in the 'general' fit
      routine. 
      relation for p[i] : an expression in reverse Polish notation, whose value
                          is to be assigned to p[i]
      Parameter p[i] is eliminated from the fit procedure, and instead equated
      to the given expression.

      Example: 
        FREL 0 p1 p2 * 3.0 -
        This results in the relation: p0 = (p1 * p2) - 3.0
For details of the syntax, give HELP RPN.

See also FREE, FIX and FLIST.

SFIT

    SFIT <data> <theo> <sim> <from> <to> <init_val0> <init_val1>

      Fits an array to a simulated spectrum.
      You have to supply:
        data                : the array containing the array to be fitted
        theo                : the array in which the final result will be put
        sim                 : the array containing the simulation
        from,to             : two integers giving the lower and upper channel
                              number of the fitting region
        init-val0,init-val1 : the initial values of the parameters
                              (2 reals giving the offset and multiplication of
                              the simulation)

VERB

    VERB <level>

      Sets the level of verbosity of mctl. The default level is 1.
      A higher level provides more detailed information.
      You have to supply:
        level : integer ranging from 0 to 3

      Affects mainly the logfile.
      To write the covariance matrix of the fitted parameters set VERB to 2.

Mössbauer tools


MCHECK

    MCHECK <unit>

      unit : a logical unit that has been associated with a Mössbauer
             data file

      The file is checked, and the runs in the file are listed.
see OPEN.

MCReate

    MCReate <name> <length> <np>

      Defines a Mössbauer spectrum. It allocates a data-space consisting
      of 4 arrays of float with dimension length, a parameter-space consisting
      of 2 arrays of double with dimension nr_par and an info-space.
      You have to supply:
        name   : spectrum name (6 characters max!)
        length : size of the data-space in channels
        np     : size of the parameter-space

      It allocates the following CIO arrays:
        name+'i'    info space
        name+'d' \
        name+'e'  \ data-space      (type 'float')
        name+'t'  /
        name+'v' /
        name+'p' \  parameter-space (type 'double')
        name+'dp'/
        name        list of all names above (CIO varlist)
see also MRM

MRM

    MRM <name>

      Deletes a Mössbauer spectrum.
      You have to supply:
        name : spectrum name (as given with MCREATE)

      All arrays listed there become undefined.
see also MCREATE.

REad

    REad <spectrum <lu> <run> <replace?>

      Reads a Mössbauer spectrum from file.
      It also initializes the error array with the square root of the data.
      A logical unit for the file should be opened beforehand, see OPEN.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
        lu       : logical unit attached to the data-file
        run      : the requested run number
        replace? : replace the current contents? (y/n)

      The following data formats may be handled
        A. Ascii files as produced by Mössbauer equipment in the nvsf group.
           Each file may consist of one or more records.
           A record has a few header lines, followed by a series of integers.
           Variations on this theme:
             mcc file:
               header line 1 is comment, preceded by a '%' in column 1
               header line 2 contains 3 integers: runnr, len, live
             dsp file:
               header line 1 contains 3 integers, 1 or 2 reals, and 1 string
                             runnr, len, live, (cal1,) cal2, unit
               header line 2 is comment
                             followed by "len" numbers in free format
             antique Mössbauer file:
               header line 1 contains 3 integers: runnr, len, live
             file of Unknown type:
               Any ascii file of "len" numbers preceded by "nskip" header lines.
               The user has to supply nskip,runnr,len from the keyboard.
           Distinction between the various ascii types is done automatically.

        B. Binary files, as made with the command PUT.

      Distinction between Ascii and Binary is made in the command OPEN.
      [In the OPEN command specify mode READ and type ASCII when the file is
      an ascii file (the usual case), or type SPECTRUM when the file is a
      binary file made with command PUT.]

FEcal

    FEcal <spectrum> <T>

      Calibrates the velocity scale using a standard iron (alpha-Fe)
      calibration measurement.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
        T        : temperature of calibration measurement

      The routine returns the (channel) position at which the expected 
      isomer shift is found. It also supplies information about the zero
      crossing.
      It fits a linear velocity calibration and calculates the velocity
      scale, as in command VK, option -1.
      (Hence it is not necessary to follow FECAL by command VK).

      FECAL assumes that the spectrum was previously folded (see FOLD)
      and then fitted (see MFIT).

FOld

    FOld <spectrum> <sum>

      Folds a Mössbauer spectrum. It reverses the second half of the
      spectrum shifts it as requested and adds the channel contents to
      those of the first half. It also recalculates the error array.
      The second half of the data- and error-array is reset.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
        sum      : an integer number

      NOTE1:
        'sum' is the average sum of the channel numbers of corresponding
        peaks in the original spectrum:
          channel[SUM - i] is added to channel[i].
      NOTE2:
        In older versions of mctl a 'shift' was asked instead of 'sum'.
        Relation:
          SHIFT = spectrum_length + 1 - SUM

INfo

    INfo <spectrum>

      Gives information about a spectrum, extracted from the spectrum's
      'info space', such as the original data file and run number.
      If this was the last spectrum fitted, it will also give the results
      of the fit.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
see also MLIST.

MOIre

    MOIre <spectrum> <gc> <fm> <AT> <PM> <first> <last> <use>

      Utility for velocity calibration using the Moire method.

        spectrum : the spectrum containing the Moire data.
        gc       : grating constant (in mu)
                   (written on the grating)
        fm       : frequency of motion (in Hz)
                   determine by counting the number of start-pulses to the MCC
                   in an interval of say 100 s
        at       : acquisition time (in s)
        pm       : steps in one period of movement (usually 512 or 1024)
        first    : first positive velocity channel
        last     : last positive velocity channel
                   Moire counts do not account for positive or negative
                   velocity the program assumes the channels between these
                   borders are positive, the rest will be multiplied by -1
        use      : 0=all,1=1st half,2=2nd half,3=fold
                   for a cosine you can use all, but for a polynomial you will
                   either have to fold the spectrum or take the halves one at
                   a time.

     The result of MOIRE is that the channels of the data-array are filled
     with velocities in mm/s. Now use FIT to get the parameters, using either
     a polynomial or a cosine, as appropriate.

VK

    VK <spectrum> <func> <vc[i]>

      Defines the velocity scale of a Mössbauer spectrum.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
        func     : -1, 0, or +1, determines the type of velocity function
          -1 :  polynomial
                vc[1]..vc[4]: velocity coefficients.
                the velocity v = vc[1] + vc[2]*k + ...
                where k = the channel number (starting with k=0)
           0 :  no velocity scale, use channel numbers as coordinates
           1 :  cosine
                vc[1] : amplitude
                vc[2] : period
                vc[3] : phase

      Note1: the velocity relation may also be set by command FECAL
      Note2: command READ does not reset the velocity relation
      Note3: MCREATE resets the velocity function to type 0
      Note4: the B,C,D parameter estimates for command MFIT are in velocity
             units, and may have to be changed after command VK

Mössbauer fit commands


DMATrix

    Dmatrix <file?>
    { <nd[n]>, ((<dmat[k][i]>,i=1,nd[n]), k=1,nlines[n]), n=1,ncomp }

      Defines the Dmatrices for the Mössbauer fit-routine.
      DMATRIX is only necessary when theories 5 or 6 of MFIT will be used.
      It should be given after command MATRIX.
      You have to supply:
        file : read from file? (y/n)
        if yes :
          lu : logical unit to read from
          id : name of the wanted matrix in the file
        if no, for each component n:
          nd[n]      : number of D-parameters in this component
        if defaults are available, no further input, otherwise:
          dmat[k][i] : nlines[n] times nd[n] matrix elements

      Defaults:
        ND[n]=Nlines[n] : all D-parameters freely variable
        ND[n]=ND[n-1]   : same D-matrix as previous component
        !!!! No defaults are used if ND[n]=Nlines[n]=2 !!!!

      Note. The layout of a Dmatrix file is slightly different:
        id (one line, 8 significant characters)
        ncomp (must be the same as for ABC matrices)
        for each component:
          nd[n], nlines[n] (must be the same as for ABC matrices)
        further data as above.

MAtrix

    MAtrix <nback> <file?> <ncomp>
    { <na[n]> <nb[n]> <nc[n]> <nlines[n]> <def[n]>, n=1..ncomp }

      Routine used to define the matrices for the Mössbauer fit-routine.
      You have to supply:
        nback : number of Background parameters (an integer)
        file? : read from file? (y/n)
          if yes:
            lu : logical unit to read from.
            id : name of the wanted matrix in the file
          if no :
            ncomp    : number of components
            repeat ncomp times:
              na     : number of A-parameters 
              nb     : number of B-parameters
              nc     : number of C-parameters
              nlines : number of lines
              def    : matrix definitions as needed

      MATRIX also releases all constraints set by MFIX or MREL.

CMATRIX

    Lists the current A, B, C and D matrices.

MFit

    MFit <spectrum> <theory> <from> <to> <init-val>

      Fit-routine of the Mössbauer-section of MCTL.
      Before MFIT, you should define the matrices (commands MAT, DMAT).
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE)
        theory   : an integer number determining the theory. Choice of
                   1 (Lorentz, amplitudes)  2 (Lorentz, areas)
                   3 (Gauss, amplitudes)    4 (Gauss, areas)
                   5 (Voigt, amplitudes)    6 (Voigt, areas)
        from,to  : two integers giving the lower and upper channel number
                   of the fitting region.
        init-val : the initial values of the parameters


    Theory 1: Lorentzian lineshape
              A=amplitude, B=position, C=FWHM
              y is background fb, plus sum over n of: - A[n]*fb*Lorentz(B,C,x)
              Lorentz(B,C,x)=1/(1 + u^2), where u=2(x-B[n])/C[n]

    Theory 2: Lorentzian lineshape
              A=peak area, B=position, C=FWHM
              y is background fb, plus sum over n of: -2/(PI*C[n])*A[n]*fb*Lorentz(B,C,x)

    Theory 3: Gaussian lineshape
              A=amplitude, B=position, C=FWHM (not sigma !!)
              y is background fb, plus sum over n of: - A[n]*fb* Gauss(B,C,x)
              Gauss(B,C,x)=exp( -0.5*u^2), where u=(x-B[n])/(C[n]/2.35)

    Theory 4: Gaussian lineshape
              A=peak area, B=position, C=FWHM (not sigma !!)
              y is background fb, plus sum over n of: -(1.06/C[n])*A[n]*fb*Gauss(B,C,x)

    Theory 5: Voigt Lorentzian × Gaussian lineshape
              A=amplitude, B=position, C=FWHM of Lorentzian, D=sigma of Gaussian
              y is background fb, plus sum over n of: - A[n]*fb*Voigt(B,C,D,x)

    Theory 6: Lorentzian × Gaussian lineshape
              A=peak area, B=position, C=FWHM of Lorentzian, D=sigma of Gaussian
              y is background fb, plus sum over n of: -2/(PI*C[n])*A[n]*fb*Voigt(B,C,D,x)

     A fit that converges too slowly may be interrupted by typing <CTRL-C>
     (press key 'CTRL' and key 'C' simultaneously)

MFIX

    MFIX <i> <j> <k> ...

      Prevents the parameters p[i],p[j],p[k],... in the Mössbauer fit
      routine, from being changed in the fit-routine.

      Note: command MATrix releases all constraints set by MFIX or MREL.
see also MFREE, MREL and MLIST

MDis

    MDis <spectrum> <from> <to> <question?> <label>

      Displays a Mössbauer spectrum.
      You have to supply:
        spectrum : a Mössbauer spectrum (as defined with MCREATE).
        from,to  : two integers giving the lower and upper channel number
                   of the display region.
        possible questions :
          theo?  : plot theoretical points     (y/n)
          meas?  : plot measured points        (y/n)
          back?  : divided by background       (y/n)
          comp?  : plot individual components  (y/n)
          extd?  : extended plot               (y/n)
        label    : plottext

      Note. If you select 'individual components', CIO arrays comp1, ...
      are also created.
see also MXPORT.

MFRee

    MFRee <i> <j> <k> ...

      Frees the parameters p[i],p[j],p[k],... in the Mössbauer fit
      routine, and releases constraints set by MFIX or MREL.
see also MFIX, MREL, MLIST, and MATRIX.

MList

    MList

      Lists the current parameters in the Mössbauer fit routine, and
      shows the correspondence between the parameters 'p' (as used in
      MFIX, MREL, and MFREE) and the actual Mössbauer parameters.
      For example, a line 'p7  (B[1])' indicates that parameter 7 corresponds
      to the first B-parameter.

      In addition MLIST list the constraints, set by MFIX and MREL.
see also MFREE, MFIX and MREL.

MRel

    MRel <i> <expression>

      Defines a constraint of the parameters used in the Mössbauer
      fit routine.
        i : parameter number
        expression : relation for parameter p[i]
                     expression in reverse Polish notation

      Parameter p[i] is eliminated from the fit procedure, and instead
      equated to the given expression.

      Example: 
        MREL 4 p1 p2 * 3.0 -
        This results in the relation:  p4 = (p1 * p2) - 3.0

      Note: command MATRIX releases all constraints set by MFIX or MREL.
For details of the syntax, give HELP RPN

see also MFREE, MFIX and MLIST.

MXport

    MXport <lu> <spectrum> <from> <to> <question?>

      Exports a fitted Mössbauer spectrum.
      You have to supply:
        lu       : a LU opened for writing (ascii mode)
        spectrum : a Mössbauer spectrum
        from,to  : two integers giving the lower and the upper channel number
                   of the region to be exported
        possible questions :
          theo?  : plot theoretical points     (y/n)
          meas?  : plot measured points        (y/n)
          back?  : divided by background       (y/n)
          comp?  : plot individual components  (y/n)

      Export fitted Mössbauer spectra. The output is suitable as input for
      program XMOS, producing high quality stacked graphs of a number of
      fitted Mössbauer spectra.

      Note. If you select 'individual components', CIO arrays comp1, ..
      are also created.
see also MDIS.


Nuclear resonant scattering fit routine


TDIS

    TDIS <spectrum> <first> <last> <question?> <label>

      Displays a nuclear resonant scattering spectrum, together with the
      results of the last TFIT command.

        spectrum    : spectrum name, as defined with MCREATE
        first, last : upper and lower channel number of the display region
        possible questions:
          theory?   : plot theoretical points     (y/n)
          measured? : plot measured points        (y/n)
          extended? : extended plot               (y/n)
        label       : a text to be written above the plot

      To save the plot, use commands PLOT and/or GEPPOST.

TFIT

    TFIT <spectrum> <theory> <first> <last>

      Nuclear Resonant Scattering Fit Routine
        spectrum   : spectrum name, as defined with MCREATE
        theory     : 10 = coherent response of hfi components
                     11 = incoherent response
        first,last : upper and lower channel number of the fit region

      Before giving command TFIT the spectrum should be defined with MCREATE
      and its data and error arrays should be filled.
      The 'velocity' array should be filled with times in ns.
      It is assumed that all time steps are equally spaced.

      Further command TFINPUT should be given to specify the layer structure.

      After TFINPUT, or in between TFIT commands, one may use TFIX, TFREE,
      TREL, and TVAR to make changes in the fit parameters and their
      constraints.

      Example of preparing the spectrum arrays for TFIT.
      If the data to be fitted are present in an ascii file in two colums,
      time and counts, respectively, then one might use the following scheme
      to input arrays spd, spe, and spv.

        open 1 <filename> read ascii
        rdas 1 2 177 spv spd   # read 177 lines spv[i], spd[i]
        move spd spe;          # copy data array to error array
        inc spe 0 176  1;      # increment by 1
        vsqrt spe spe;         # set error = sqrt(data + 1)

      A fit that converges too slowly may be interrupted by typing <CTRL-C> 
      (press key 'CTRL' and key 'C' simultaneously).

TFINPUT

    Input for nuclear resonant scattering fit routine.
    Should be called before command TFIT.

    In the following 'variation' is the amount by which to vary a parameter
    for the calculation of derivatives in the fit procedure.
    If a variation is given the value 0, the corresponding parameter is
    kept fixed in the fit.
    Parameter constraints may also be given by TVAR, TFIX, TREL, and TFREE.

    You have to supply:
      spectrum : name of a spectrum structure as defined with MCREATE
      id       : a line of text
      theta    : grazing incidence angle (mrad)
      back     : estimate of background
      time     : estimate of time shift (ns)
      scale    : scaling factor
      nblock   : number of blocks (repetive structures)
      For each block:
        nrep   : the number of times this block is repeated
        nlayer : the number of layers in this block
        For each layer:
          thickness, variation in thickness (nm)
          roughness, variation in roughness (nm)
          nel  : number of elements in this layer
          For each element:
            z          : atomic number
            n          : atomic density (atoms/nm3)
            f1,f2      :
          nhfi : number of hfi components in this layer
          For each hfi component:
            isomer shift, variation in isomer shift (Mrad/s)
            hf field, variation in hf field (T)
            quadrupole coupling, variation in quadrupole coupling (Mrad/s)
            width sigma, variation in sigma (Mrad/s)
            recoilless fraction, variation in recoilless fraction
            weigth, variation in weight
            m=0 intensity ksi, variation in ksi

TFIX

    TFIX <i> <j> <k> ...

      Prevents the parameters p[i],p[j],p[k],... in the fit routine TFIT
      from being changed in the fit.

      Note: command TFINPUT releases all constraints set by TFIX or TREL,
      but fixes all parameters for which a variation of 0 is given.
see also TFREE, TREL and TLIST.

TFREE

    TFREE <i> <j> <k> ...

      Frees the parameters p[i],p[j],p[k],... in the fit routine TFIT,
      and releases constraints set by TFIX or TREL.
      Take care not to free a parameter for which variation was set to 0
      by TFINPUT or TVAR.

TLIST

    Lists the constraints, set by TFINPUT, TVAR, TFIX, and TREL.

TREL

    TREL <i> <expression>

      Defines a constraint of parameter p[i] used in the TFIT command.
      Parameter p[i] is eliminated from the fit procedure, and instead
      equated to the given expression.
      If the parameter was previously fixed (by TFINPUT, TVAR or TFIX)
      a warning is given.
For further details in the syntax, give HELP RPN.

TVAR

    TVAR <i> <new_value> <new_variation>

      May be used to vary the current value and variation of
      parameter p[i] in between commands TFIT. 
      If the variation is set to 0, and the parameter was free,
      then it is now fixed, and vice versa. However if a parameter
      was eliminated (by TREL) it is not automatically fixed or freed.

» back to old software