[Previous]
[Next]
NAME:
cBit32
PURPOSE:
Put the binary representation for a integer*4 into a 32 character string array.
CALLING SEQUENCE:
character*32 function cBit32(IN)
INPUTS:
IN integer*4 integer to be converted to/from binary
CALLS: ***
Bit32
RESTRICTIONS:
If IN is negative, twos complement convention is assumed, i.e. bit 16 is set
to one; the other 15 bits are those of the positive number IN+32768.
PROCEDURE:
The least significant bit of IN is put into cBit32(32:32), the most
significant in cBit32{1:1).
MODIFICATION HISTORY:
MAR-1992, Paul Hick (UCSD/CASS; pphick@ucsd.edu) : Rewrite of JBIT16
[Previous]
[Next]
NAME:
cDbl2Str
PURPOSE:
Convert floating point number to string
CATEGORY:
Strings: write floating point number to string
CALLING SEQUENCE:
character*14 function cDbl2Str(Fin,Nin)
INPUTS:
Fin double precision floating point number to be converted
Nin integer (see PROCEDURE)
OUTPUTS:
cDbl2Str character*(*) resulting string
CALLS: ***
Dbl2Str
CALLED BY:
smei_base, smei_cal, smei_orb, smei_skyd
SEE ALSO:
Dbl2Str
RESTRICTIONS:
cDbl2Str is a character*14 function
PROCEDURE:
See Flt2Str
MODIFICATION HISTORY:
AUG-1995, Paul Hick (UCSD)
[Previous]
[Next]
NAME:
cFilePath
PURPOSE:
Platform-independent construction of file name
CATEGORY:
Portability
CALLING SEQUENCE:
character*256 function cFilePath(cRootDir,nSub,cSub,cFile)
CALLS: ***
iFilePath
PROCEDURE:
The length of the function is set equal to the value of
parameter FIL__LENGTH in filparts.h
MODIFICATION HISTORY:
JUN-2004, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cFlt2Str
PURPOSE:
Convert floating point number to string
CATEGORY:
Strings: write floating point number to string
CALLING SEQUENCE:
character*14 function cFlt2Str(Fin,Nin)
INPUTS:
Fin real floating point number to be converted
Nin integer (see PROCEDURE)
OUTPUTS:
cFlt2Str character*(*) resulting string
CALLS: ***
Flt2Str
CALLED BY:
Connect, HOSOrbID, HOSPlot, LOSReach, LOSWeights, Pandora, PrintAll, WR2DARR
XMAP_SC_POS, XMAP_SC_POS8, nrRatInt, smei_frm_base, smei_frm_c3fudge, smei_frm_ped
smei_frm_ped_guess, smei_frm_ratio
SEE ALSO:
Flt2Str
RESTRICTIONS:
cFlt2Str is a character*14 function
PROCEDURE:
See Flt2Str
MODIFICATION HISTORY:
AUG-1995, Paul Hick (UCSD)
[Previous]
[Next]
NAME:
Chebyshev
PURPOSE:
Evaluate value of Chebyshev polynomial in given point X
CATEGORY:
Math: orthogonal polynomials
CALLING SEQUENCE:
function Chebyshev(X,N)
INPUTS:
X real the Chebyshev polynomial is evaluated in X
N integer degree of Chebyshev polynomial
OUTPUTS:
Chebyshev real value of n-th degree Chebyshev polynomial in XS
RESTRICTIONS:
The degree N must be larger/equal zero
PROCEDURE:
Evaluation is based three-term recurrence relation for Chebyshev
polynomials (T_0 = 1 ; T_1 = x)
T_i = 2xT_i-1 - T_i-2 ; i = 2,3,4, ...
MODIFICATION HISTORY:
DEC-1991; Paul Hick (UCSD)
[Previous]
[Next]
NAME:
Chebyshev_exp
PURPOSE:
Given the coefficients of its expansion in Chebyshev polynomials,
evaluate the value of a polynomial in point X
CATEGORY:
Math: orthogonal polynomials
CALLING SEQUENCE:
function Chebyshev_exp(A,B,X,N,CN)
INPUTS:
A,B real definition interval (see PROCEDURE)
X real X-value where function is evaluated
N integer # terms in Chebyshev expansion
CN(N) real expansion coefficients
OUTPUTS:
Chebyshev_exp real value of polynomial in point X
RESTRICTIONS:
> The number of terms in the expansion, N, must be larger then zero
> A and B must be unequal
PROCEDURE:
> The Chebyshev expansion is
VALUE(X) = SUM(i=1,N) { CN(I)*T_i-1(XS) }, i.e. CN(I) is the
coefficient of the Tchebyshev polynomial of degree I-1
> The interval [a,b] is mapped to [-1,1]. X is rescaled accordingly to
XS = (2X-(B+A))/(B-A)
> The evaluation of the Tchebyshev polynomials is based on the
three-term recurrence relation (T_0 = 1 ; T_1 = x)
T_i = 2xT_i-1 - T_i-2 ; i = 2,3,4, ...
MODIFICATION HISTORY:
DEC-1991; Paul Hick (UCSD)
[Previous]
[Next]
NAME:
CheckMass
CALLING SEQUENCE:
subroutine CheckMass(XCbeg,XCend,V3D,D3D)
CALLS: ***
Flt2Str, Say, Str2Str, T3D_get_grid, T3D_iget, cosd, sind
CALLED BY:
SW_Model_Kinematic
INCLUDE:
include 'math.h'
include 't3d_array.h'
include 't3d_grid_fnc.h'
include 't3d_loc_fnc.h'
[Previous]
[Next]
NAME:
cHideLogical
PURPOSE:
Hides the value of a logical from a fully qualified file specification
CATEGORY:
String handling
CALLING SEQUENCE:
character*256 function cHideLogical(cStr)
INPUTS:
cStr character*(*) Fully-qualified file name
OUTPUTS:
cHideLogical character*(*) File name with leading part replaced
by the name of a logical.
CALLS: ***
iHideLogical
CALLED BY:
BField_Choose, BField_Get, ReadVIPS, ReadVIPSLOSCheck, ReadVIPSLOSCheckn
ReadVIPSn, dailyips [1], dailyips [2], iSearch, smei_base, smei_orb
RESTRICTIONS:
Probably does not work on VMS
PROCEDURE:
If the input file is /home/soft/smei/test.txt, and
logical $smei is defined as /home/soft/smei then the output
string will be $smei/test.txt.
The length of the function is set equal to the value of
parameter FIL__LENGTH in filparts.h
MODIFICATION HISTORY:
JUN-2004, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cInt2Str
PURPOSE:
Write integer into string subject to constraints set by Int2StrSet
CATEGORY:
Strings: writing integer to string
CALLING SEQUENCE:
character*14 function cInt2Str(I) ! Same length as cFlt2Str
INPUTS:
I integer integer to be processed
OUTPUTS:
cInt2Str character*(*) output string
CALLS: ***
Int2Str
CALLED BY:
Connect, ForeignArg, HOSOrbID, HOSPlot, HOSPlot_bCommand, HOSRead, HOSUpdate, HOSWrite
HOSdos2vms, LOSReach, NicHdr, OGetRecord, Pandora, SD, T3D_Read_B, dailyips [1]
dailyips [2], iHOSInfo, iHOSRead, iHOSWrite, iOpenFile, rice, say_fts, smei_cal_c3mask
smei_cal_get, smei_cal_init, smei_cal_read, smei_frm_path, smei_orb, smei_skyd
smei_skyd_combine, smei_skyd_fill, smei_skyd_flush, smei_skyd_fts
smei_skyd_init, smei_skyd_make, smei_skyd_pixel, smei_skyd_sky, smei_skyd_sort
SEE ALSO:
cFlt2Str
PROCEDURE:
cInt2Str is declared as a character*14 function
MODIFICATION HISTORY:
JAN-1995, Paul Hick (UCSD)
[Previous]
[Next]
NAME:
Clean0
PURPOSE:
Cleans 'dirty' or 'NaN' REAL*4 zeros
CATEGORY:
gen/for/lib
CALLING SEQUENCE:
subroutine Clean0 (idirty)
INPUTS:
X real value to be checked
OUTPUTS:
X real set to zero if it was a dirty zero
I integer 0: if the was not modified
1: if the input was set to zero
CALLS:
(none)
SEE ALSO:
RClean, iClean0, iCleanNaN
RESTRICTIONS:
Should be called with REAL*4 argument!
PROCEDURE:
A dirty zero is tested by the condition
iand(idirty,Z'0000FF80') .eq. 0
The NaN value is tested by
iand(idirty,Z'0000FF80') .eq. 0
MODIFICATION HISTORY:
SEP-1999, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
Connect
PURPOSE:
Connects separate data files into one big file
CATEGORY:
I/O: Helios
CALLING SEQUENCE:
program Connect
INPUTS:
> The input files contain records in the usual structure (PHOTO_DATA or
90DEG_DATA). They should contain only one particular color, but may
contain any number of photometers and filters.
> File names, start-, end- and transfer times (see procedure)
are read either from a file cFileL or from the keyboard.
> The file cFileL starts with an optional record containing two times
TBeg and TEnd indicating that only data between these two time limits
should be extracted. If this record is omitted TBeg=0 and TEnd=999.
Each subsequent record contains a file name and the corresponding
transfer time TNew, separated by a blank.
> All times are in day of year
OUTPUTS:
The output file name OUT (recl=37) is a 13 character string 'XYRLF_doy.ext'
X = A or B for HELIOS A or B respectively;
YR = the year of observation (i.e. 78 for 1978);
L = U,B,V for U/B/V light respectively;
F = 1 through 5 depending on the filter (if the file contains
15/30 deg photometer data); or
= 9 if the file contains only 90 deg photometer data (see NOTE);
doy = the day of year for the earliest data point.
ext = ZLD for unnormalized data files (i.e. original zodiacal light
data); or
= DAT for normalized data files
The 90 deg data are optionally written to either a separate file
(with a '9' as fifth character and recl 6) or are appended to the file
OUT.
CALLS: ***
ArrI4Index, ArrI4Zero, ArrR4Copy, AskR4, AskWhat, AskYN, IndexR4, Int2Str, Say, Str2Str
Str2StrSet, bOpenFile, cFlt2Str, cInt2Str, iFilePath, iFreeLun, iGetFileSpec, iHOSInfo
iHOSRead, iHOSWrite, iPutFileSpec, iSetFileSpec, iUniqueName, iwhitespace
INCLUDE:
include 'dirspec.h'
include 'filparts.h'
include 'openfile.h'
include 'str2str_inc.h'
include 'hos_e9.h'
RESTRICTIONS:
> A zodiacal light model may already be subtracted ('normalized' data),
but not necesarily. All files should be of the same type: all or none
of the files are normalized; all files should contain the same color;
if this is not the case the program stops and no output is produced.
> WARNING: if you want to connect data from different years, say from
december 1976 to january 1977, 366 days should be added to the times
for the later year, e.g. doy 1.2345 of 1977 should become 367.2345.
There is an option in PANDORA which does that.
PROCEDURE:
> One specified filter will be extracted. The output file is sorted in
time. (Note that the program can also be run on one file, to extract
data for a specific time period, or to extract a specific filter).
> A period of time is covered by N subsequent (partly overlapping) data
files, file I+1 containing data later than file I (I=1,..,N-1).
Data will be copied to the output file using the following rules:
(1) Only data inside an overall time window TBeg <= T <= TEnd are
extracted
(2) The switch from one data file to the next is made at transfer times
TNew, which have to be specified for each file except the last
(if a value is given it should be larger than 999).
> For the I-th file data will be copied for which
TNew(I-1) <= T < TNew(I),
where TNew(0) = TBeg and TNew(N) = TEnd;
i.e. data up to the transfer time are copied; for the next file data
from the transfer time onwards are copied.
> The output files are created in the directory where the first data file
is found (usually, but not necessarily the working directory).
MODIFICATION HISTORY:
1989, Paul Hick (UCSD)
[Previous]
[Next]
NAME:
ConvertG2D
PURPOSE:
Make a density map at a given height from a G-level map from the
Hick et al. map technique.
CATEGORY:
Data processing
CALLING SEQUENCE:
subroutine ConvertG2D(ID,N,G,D,RR,D1AU,PW)
INPUTS:
ID integer mode - 0=standard, 1=input conversion
N integer # data points
G(N) real G-levels
RR real Reference distance (AU)
D1AU real 1 AU density
PW real Power
OUTPUTS:
D(N) real Density map at height RR
CALLS: ***
BadR4, GTODEN
CALLED BY:
ipsd, ipsdt
PROCEDURE:
Values in G = BadR4() are not converted (D is set to BadR4()).
iGfnc = 0 (n=G)
iGfnc = 1 (Tappin, n=9xG^2)
iGfnc = 2 (n=10.4xG^3.4)
iGfnc = 3 (n=9.5xG^5.1)
MODIFICATION HISTORY:
NOV, 1995 B. Jackson (STEL,UCSD)
[Previous]
[Next]
NAME:
CopyDtoDVN
PURPOSE:
Make density maps at the velocity map time cadence. Velocity and density maps of the same latitude and longitude
grid spacing must be input. The Carrington rotation beginning and endings at each time are imput, but can change
with time as long as their interval lengths remain constant. Thus, either a corotating (tested) or an inertial
set of maps can be converted.
If Mode = 0, then only an interpolated value of the density map is transferred.
If Mode = 1, then an interpolated value of the density map is transferred plus an interpolated map out to many
sigma in time at that longitude.
If Mode = 2, then the interpolated value is cut off at the 1.5 times the velocity temporal interval -
i.e. one time from the non-valid value in time.
CATEGORY:
Data processing
CALLING SEQUENCE:
call CopyDtoDVN(Mode,XCbe,XCtbeg,XCtend,XCtbegG,XCtendG,nLng,nLat,nTG,nTV,ConsT,DD,DDV,DDT,DDD)
INPUTS:
Mode integer 0 - transfer good time values only
1 - transfer both good and average time values
2 - Same as 1, but cut off filter after 1 time away
XCbe real Beginning V times (XCbeV)
XCtbeg real*8 Beginning V time (XCtbegV)
XCtend real*8 Ending V time (XCtendV)
XCtbegG real*8 Beginning G time
XCtendG real*8 Ending G time
nLng integer # Longitude points
nLat integer # Latitude points
nTG integer # G-level times
nTV integer # velocity times
ConsT real Time filter constant (in terms of the velocity time interval)
DD(nLng,nLat,nTG) real Density map
OUTPUTS:
DDV(nLng,nLat,nTV) real Density map interpolated at the velocity time cadence
SCRATCH ARRAYS:
DDT(nTG) real
DDD(nLng) real
CALLS: ***
BadR4, FLINT
CALLED BY:
ipsdt
PROCEDURE:
Bad values (indicated by BadR4()) are not processed in Mode 0
MODIFICATION HISTORY:
NOV, 1999 B. Jackson (STEL,UCSD)
[Previous]
[Next]
NAME:
CopyVtoVDN
PURPOSE:
Make velocity maps at the density map time cadence. Velocity and density maps of the same latitude and longitude
grid spacing must be input. The Carrington rotation beginning and endings at each time are imput, but can change
with time as long as their interval lengths remain constant. Thus, either a corotating (tested) or an inertial
set of maps can be converted.
If Mode = 0, then only an interpolated value of the velocity map is transferred.
If Mode = 1, then an interpolated value of the velocity map is transferred plus an interpolated map out to many
sigma in time at that longitude.
If Mode = 2, then the interpolated value is cut off at the 1.5 times the density temporal interval -
i.e. one time from the non-valid value in time.
CATEGORY:
Data processing
CALLING SEQUENCE:
call CopyVtoVDN(Mode,XCbe,XCtbeg,XCtend,XCtbegV,XCtendV,nLng,nLat,nTV,nTG,ConsT,VV,VVD,VVT,VVV)
INPUTS:
Mode integer 0 - transfer good time values only
1 - transfer both good and average time values
2 - Same as 1, but cut off filter after 1 time away
XCbe real Beginning G (XCbeGG)
XCtbeg real Beginning G time (XCtbegG)
XCtend real Ending G time (XCtendG)
XCtbegV real Beginning V time (XCtbegV)
XCtendV real Ending V time (XCtendV)
nLng integer # Longitude points
nLat integer # Latitude points
nTV integer # velocity times
nTG integer # G-level times
ConsT real Time filter constant (in terms of the density time interval)
VV(nLng,nLat,nTV) real Velocity map
OUTPUTS:
VVD(nLng,nLat,nTG) real Density velocity map
SCRATCH ARRAYS:
VVT(nTV) real
VVV(nLng) real
CALLS: ***
BadR4, FLINT
CALLED BY:
ipsdt
PROCEDURE:
Bad values (indicated by BadR4()) are not processed in Mode 0
MODIFICATION HISTORY:
NOV, 1999 B. Jackson (STEL,UCSD) Found error 11/14/00 in determining first part BVJ
[Previous]
[Next]
NAME:
cosd
PURPOSE:
Goniometric function accepting arguments in degrees
CATEGORY:
Math
CALLING SEQUENCE:
function cosd(a)
INPUTS:
a,b real angle in degrees
c,d double precision angle in degrees
OUTPUTS:
S real angle in degrees
D double precision angle in degrees
CALLED BY:
AsymDust, CheckMass, CosdLookup, DustAsymmetry, ElSunDistance, HOSOrbID, KeplerOrbit
MkDMap, MkDMaptdn, MkLOSWeights, MkPos, MkPostd, MkVMap, MkVMaptdN, PA_POLE, PointPOnLOS
PointPOnLOS_Near, PointPOnLOS_Reset, ReadG, SC_ECLIP, SC_ECLIP90, SphereWeight
THOM_WTF, ThomsonLOS, ThomsonLOS3D, ThomsonLOS3DStep, ThomsonMidpoint
ThomsonMidpointFar, Time2PrecesssionLtd, Write3D_bb, Write3D_bb_UT
Write3D_bb_XC, Write3D_bbtt, Write3D_nv_UT, iProcessOoty, iProcessOotyn
iProcessUCSD, iReadOoty, iReadOotyn, iReadUCSD
INCLUDE:
include 'math.h'
SEE ALSO:
acosd, asind, atan2d, atand, dacosd, dasind, datan2d, datand, dcosd, dsind, dtand, sind, tand
RESTRICTIONS:
Note that it probably will be necessary to declare the function
type explicitly in the calling program, e.g.
real acosd, cos
double precision dacosd, dcos
PROCEDURE:
Can be used with Fortran compilers that do not provide goniometric
function accepting input and providing output in degrees (e.g. g77).
MODIFICATION HISTORY:
JAN-2001, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
CosdLookup
PURPOSE:
Calculates cos(x) using lookup table
CATEGORY:
Math
CALLING SEQUENCE:
function CosdLookup(X)
INPUTS:
X real X-coordinate
OUTPUTS:
F real cos(x) from table lookup
CALLS: ***
FLINT, cosd
SEE ALSO:
GaussLookup, SinLookup
MODIFICATION HISTORY:
SEP-2003, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cotri
PURPOSE:
include file used by HELIM program
[Previous]
[Next]
name: CRX_ped_dark_cam1.for
purpose: determines pedestal & dark current, finds measles
inputs: *.nic data files listed in "namelist2.txt"
outputs: before and after *.grd files
" Note here that the pedestal, dark, and CR information incorporated into the "trailer"
" The output files xxx.dat,txt present information for graphs/subseq of the various quantities
calls: subroutines measles (+ plane), and INDEXX (2 versions)
mods: April 2004, Andrew Buffington (UCSD; (858)-534-6630; abuffington@ucsd.edu)
[Previous]
[Next]
name: CRX_ped_dark_cam2.for
purpose: determines pedestal & dark current, finds measles
inputs: *.nic data files listed in "namelist2.txt"
outputs: before and after *.grd files
" Note here that the pedestal, dark, and CR information incorporated into the "trailer"
" The output files xxx.dat,txt present information for graphs/subseq of the various quantities
calls: subroutines measles (+ plane), and INDEXX (2 versions)
mods: April 2004, Andrew Buffington (UCSD; (858)-534-6630; abuffington@ucsd.edu)
[Previous]
[Next]
NAME:
CRX_ped_dark_cam3.for
PURPOSE:
determines pedestal & dark current, finds black/white measles & cosmic rays
INPUTS:
*.nic data files listed in "namelistCRX.txt"
OUTPUTS:
before and after *.grd files
Note here that the pedestal, dark, and CR information incorporated into the "trailer"
The output file xxx.dat presents information for graphs of the various quantities
CALLS:
and INDEXX, subroutines measles (+ plane)
MODIFICATION HISTORY:
January 2004, Andrew Buffington (UCSD; (858)-534-6630; abuffington@ucsd.edu)
Please Note that "headroom" and small-scale FF are turned OFF when camera 3 in 2x2 mode as here...
Another consequence of this is that columns 11 and 631 are NOT half stamped out as would be
the case if the SSFF were imposed. Thus these have a full amount of pedestal and dark current,
but only an unknown and variable-with-row fraction of sky between 0.5 and unity. Yuck!
[Previous]
[Next]
NAME:
cTime2Str
PURPOSE:
Convert time to specified string format
CATEGORY:
gen/for/lib
CALLING SEQUENCE:
character*80 function cTime2Str(cFmt,t)
OUTPUTS:
u(2) integer rounded 2-element standard time
CALLS: ***
Time2Str
CALLED BY:
jpl_message, smei_cal_get, smei_cal_init, smei_frm_c3fudge, smei_frm_ratio
MODIFICATION HISTORY:
JAN-2004, Paul Hick (pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cTime2System
PURPOSE:
Get system time in specified string format
CATEGORY:
gen/for/lib
CALLING SEQUENCE:
character*80 function cTime2System(cFmt)
OUTPUTS:
u(2) integer rounded 2-element standard time
CALLS: ***
Time2Str, Time2System
CALLED BY:
ExtractInsitu, Extractd, Extractd3d, Say, StopWatch, WR2DARR, iUniqueName, sprint
MODIFICATION HISTORY:
JAN-2004, Paul Hick (pphick@ucsd.edu)
[Previous]
[Next]
NAME:
CvD2G
PURPOSE:
Convert between normalized density nr^2 and g^2
CATEGORY:
Tomography
CALLING SEQUENCE:
subroutine CvD2G(PWN,N,DD,G2)
INPUTS:
PWN real
N integer # values to be converted
N < 0 : check for bad values (BadR4())
DD(N) real normalized densities
OUTPUTS:
G2(N) real g^2 values
CALLS: ***
BadR4
CALLED BY:
SW_Model_Kinematic, ipsg2, ipsg2s, ipsg2t
PROCEDURE:
The relation between g^2 and normalized density:
G2 = DD^(2*PWN)
If N < then bad values in the output array will also be
bad in the output array.
MODIFICATION HISTORY:
MAY-1999, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cvdim
PURPOSE:
Include file used by contour plotting routines CVPLOT and CVYN
[Previous]
[Next]
NAME:
CvG2D
PURPOSE:
Convert from g^2 to normalized density nr^2
CATEGORY:
Tomography
CALLING SEQUENCE:
subroutine CvG2D(PWN,N,G2,DD)
INPUTS:
PWN real
N integer # values to be converted
N < 0 : check for bad values (BadR4())
G2(N) real g^2 values
OUTPUTS:
DD(N) real normalized densities
CALLS: ***
BadR4
CALLED BY:
SW_Model_Kinematic, Write3D_nv_UT, Write3D_nv_XC, ipsg2, ipsg2s, ipsg2t
PROCEDURE:
The relation between g^2 and normalized density:
G2 = DD^(2*PWN)
If N < then bad values in the output array will also be
bad in the output array.
MODIFICATION HISTORY:
MAY-1999, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
CvI2
PURPOSE:
Convert integer*2 numbers between VMS, DOS & UNIX
CALLING SEQUENCE:
subroutine CvI2(cOp,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
N integer # elements in XX
XX(N) integer*2 array
OUTPUTS:
XX(N) integer*2 converted array
CALLED BY:
iHOSRead, iHOSWrite
RESTRICTIONS:
cOp should correspond to an entry in the include file 'dirspec.h'.
INCLUDE:
include 'dirspec.h'
CALLS: ***
CvSwap
PROCEDURE:
Byte swapping
[Previous]
[Next]
NAME:
CvI4
PURPOSE:
Convert integer*4 numbers between VMS, DOS & UNIX
CALLING SEQUENCE:
subroutine CvI4(cOp,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
N integer # elements in XX
XX(N) integer*4 array
OUTPUTS:
XX(N) integer*4 converted array
CALLED BY:
jpl_close, jpl_init, jpl_state
RESTRICTIONS:
cOp should correspond to an entry in the include file 'dirspec.h'.
INCLUDE:
include 'dirspec.h'
CALLS: ***
CvSwap
PROCEDURE:
Byte swapping
[Previous]
[Next]
NAME:
cvntvms
CALLS: ***
CvR4S, CvR8S
[Previous]
[Next]
NAME:
CvR4
PURPOSE:
Converts real*4 numbers between VMS, DOS and Unix
CATEGORY:
gen/for/lib
CALLING SEQUENCE:
subroutine CvR4(cOp,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
N integer # elements in XX
XX(N) real*4 array
OUTPUTS:
XX(N) real*4 converted array
CALLS: ***
CvSwap
CALLED BY:
HERDISK, iHOSArch, iHOSRead, iHOSWrite
RESTRICTIONS:
cOp should correspond to an entry in the include file 'dirspec.h'.
Supported are 'DOS' (DOS/WIN/NT), 'UNX' (unix), 'VMS' and 'LNX' (Linux).
INCLUDE:
include 'dirspec.h'
include 'math.h'
PROCEDURE:
Usually the numbers are converted to the native OS (i.e. the OS specified
in 'dirspec.h'), and only the OS of origin needs to specified.
CvR4S and CvR8S allow specification of both origin and destination OS
for the conversion.
(This setup is not that great. Number representation is more a matter of
hardware architecture than OS. E.g. Linux and DOS/Windows running on Intel
or AMD use the same number representations. So we should be testing for
CPU architecture rather than OS. This is left as an excersize to the reader,
I suppose).
INTEGER CONVERSIONS:
VMS and DOS use the same representation for integers.
Unix has the bytes in the opposite order relative to VMS and DOS (I think)
Linux is the same as DOS.
REAL CONVERSIONS:
The only difference between DOS and Unix appears to be that the order of the
bytes is reversed (as for integers).
Linux is the same as DOS.
The real work is the conversion between DOS and VMS and v.v.
Real*8 structure (64 bits):
<------------------------ fraction -----------------> S<--- Exp --><--> VMS
S<--- Exp --><-------------------- fraction --------------------------> INTEL
32109876 54321098 76543210 98765432 10987654 32109876 54321098 76543210
Real*4 structure (32 bits):
<--- fraction --> S<- Exp -><-----> VMS
S<- Exp -><--- fraction ----------> INTEL
10987654 32109876 54321098 76543210
For both real*4 and real*8 the position of exponent and fraction in VMS and DOS
are matched by reversing the order of the 2-byte words.
Real*8 (G-floating) numbers originating on the VAX have the following structure:
Bit 0..3,16..63 : normalized 52-bit fraction; the 53th bit (J-bit) is omitted, and is
always 1, i.e. the numbers are always normalized
Bit 4..14 : 11-bit biased exponent (biasing constant is 1025)
Bit 64 : sign bit
Real*8 for the INTEL FPU have the following structure
Bit 0..51 : normalized 52-bit fraction; the 53th bit (J-bit) is omitted, but contrary to
the VAX it can be zero (for very small, 'denormalized', numbers).
Bit 52..62 : 11-bit biased exponent (biasing constant is 1023)
Bit 63 : sign bit
Real*4 (F-floating) numbers originating on the VAX have the following structure:
Bit 0..6,16..31 : normalized 24-bit fraction; the 24th bit (J-bit) is omitted, and is
always 1, i.e. the numbers are always normalized
Bit 7..14 : 8-bit biased exponent (biasing constant is 129)
Bit 15 : sign bit
As far as I can tell the exponent is always non-zero (1<=exp<=255), except for the
number zero (when all bits are zero). An exponent of zero combined with a non-zero fraction
are not valid real numbers.
Real*4 for the INTEL FPU have the following structure
Bit 0..22 : normalized 24-bit fraction; the 24th bit (J-bit) is omitted, but contrary to
the VAX it can be zero (for very small, 'denormalized', numbers).
Bit 23..30 : 8-bit biased exponent (biasing constant is 127)
Bit 31 : sign bit
VAX -> INTEL:
The VAX structure is mapped to the INTEL structure by reversing the order of the
words (2 for real*4, 4 for real*8.
The VAX exponent is usually two more than the INTEL exponent (the difference of the
biasing constants). If after subtraction of 2 the exponent is still positive, the
subtraction completes the conversion from a VAX to an INTEL real*4 or real*8. If the
corrected exponent is 0 or -1 than the INTEL real is denormalized and the fraction
must be updated also. This is done by shifting the bits in the fraction downward
(decreasing). Note that this means that the J-bit is shifted into the 'visible' 23-
or 52-bit part of the fraction.
INTEL -> VMS:
The Intel architecture uses exponent 255 (real*4) and 2047 (real*8) for special
purposes: quiet NaN, signal NaN, plus and minus infinity. These are not available
on VMS. These values will be set to the parameter values MATH__NARN or MATH__NARN8
and +/-MATH__PINF or +/-MATH__PINF8 defined in include file math.h.
Numbers with exponent 254 (real*4) and 2046 (real*8) are outside the range of
real values covered by VMS, and are set to +/-MATH__PINF or +/-MATH__PINF8.
Numbers with exponent in the range [1,253] (real*4) and [1,2045] (normalized numbers)
are converted to VMS by adding 2 to the exponent.
Numbers with exponent 0 (denormalized numbers) can be normalized on the VMS only
if the fraction can be made >=1 by shifting bits at most 2 positions. If it takes
more than 2 the number will be set to zero.
MODIFICATION HISTORY:
OCT-1998, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
TEST PROGRAM
program test
include 'dirspec.h'
include 'openfile.h'
include 'math.h'
character cOp*3
logical bWrite
real*4 xx
real*8 dd
integer i4(2)
equivalence (dd,i4),(xx,i4)
real*4 xdos
real*8 ddos
integer idos(2)
equivalence (ddos,idos),(xdos,idos)
real*4 xvms
real*8 dvms
integer ivms(2)
equivalence (dvms,ivms),(xvms,ivms)
dd = MATH__NARN8
cOp = OS__VMS
if (cOpSys .eq. OS__VMS) cOp = OS__DOS
xx = .29E-36
I0 = I
do while (I-I0 .lt. 60)
I = I+1
ii = 0
if (cOpSys .eq. OS__DOS) call MVBITS(i4(1),23,8,ii,0)
if (cOpSys .eq. OS__VMS) call MVBITS(i4(1), 7,8,ii,0)
xvms = xx
call CvR4S(cOpSys,cOp,1,xvms)
xdos = xvms
call CvR4S(cOp,cOpSys,1,xdos)
print *, ' exp ',ii, xx, xdos, xdos-xx, idos(1)-i4(1)
xx = xx/1.1
end do
xx = MATH__PINF/100.
I0 = I
do while (I-I0 .lt. 10)
I = I+1
ii = 0
if (cOpSys .eq. OS__DOS) call MVBITS(i4(1),23,8,ii,0)
if (cOpSys .eq. OS__VMS) call MVBITS(i4(1), 7,8,ii,0)
xvms = xx
call CvR4S(cOpSys,cOp,1,xvms)
xdos = xvms
call CvR4S(cOp,cOpSys,1,xdos)
print *, ' exp ',ii, xx, xdos, xdos-xx, idos(1)-i4(1)
if (cOpSys .ne. OS__VMS .or. xx .lt. MATH__PINF/2.) xx = xx*2
end do
dd = 0.56D-306
I0 = I
do while (I-I0 .lt. 60)
I = I+1
ii = 0
if (cOpSys .eq. OS__DOS) call MVBITS(i4(2),20,11,ii,0)
if (cOpSys .eq. OS__VMS) call MVBITS(i4(1), 4,11,ii,0)
dvms = dd
call CvR8S(cOpSys,cOp,1,dvms)
ddos = dvms
call CvR8S(cOp,cOpSys,1,ddos)
print *, ' exp ',ii, dd, ddos, ddos-dd, idos(1)-i4(1), idos(2)-i4(2)
dd = dd/1.1d0
end do
dd = MATH__PINF8/100.
I0 = I
do while (I-I0 .lt. 10)
I = I+1
ii = 0
if (cOpSys .eq. OS__DOS) call MVBITS(i4(2),20,11,ii,0)
if (cOpSys .eq. OS__VMS) call MVBITS(i4(1), 4,11,ii,0)
dvms = dd
call CvR8S(cOpSys,cOp,1,dvms)
ddos = dvms
call CvR8S(cOp,cOpSys,1,ddos)
print *, ' exp ',ii, dd, ddos, ddos-dd, idos(1)-i4(1), idos(2)-i4(2)
if (cOpSys .ne. OS__VMS .or. dd .lt. MATH__PINF8/2) dd = dd*2d0
end do
end
[Previous]
[Next]
NAME:
CvR4S
PURPOSE:
Convert real*4 numbers between VMS, DOS & UNIX
CALLING SEQUENCE:
entry CvR4S(cOp,cOpDest,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
cOpDest character*3 operating system to be converted to
N integer # elements in XX
XX(N) real*4 array
OUTPUTS:
XX(N) real*4 converted array
CALLED BY:
cvntvms
RESTRICTIONS:
cOp and cOpDest should correspond to an entry in the include file 'dirspec.h'.
CALLS: ***
CvSwap
PROCEDURE:
See CvR4
[Previous]
[Next]
NAME:
CvR8
PURPOSE:
Convert real*8 numbers between VMS, DOS & UNIX
CALLING SEQUENCE:
subroutine CvR8(cOp,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
N integer # elements in XX
XX(N) real*8 array
OUTPUTS:
XX(N) real*8 converted array
CALLED BY:
jpl_close, jpl_init, jpl_state
RESTRICTIONS:
cOp should correspond to an entry in the include file 'dirspec.h'.
INCLUDE:
include 'dirspec.h'
include 'math.h'
CALLS: ***
CvSwap
PROCEDURE:
See CvR4
[Previous]
[Next]
NAME:
CvR8S
PURPOSE:
Convert real*8 numbers between VMS, DOS & UNIX
CALLING SEQUENCE:
entry CvR8S(cOp,cOpDest,N,XX)
INPUTS:
cOp character*3 operating system where the numbers originated
cOpDest character*3 operating system to be converted to
N integer # elements in XX
XX(N) real*8 array
OUTPUTS:
XX(N) real*8 converted array
CALLED BY:
cvntvms
RESTRICTIONS:
cOp and cOpDest should correspond to an entry in the include file 'dirspec.h'.
CALLS: ***
CvSwap
PROCEDURE:
See CvR4
[Previous]
[Next]
NAME:
CvSwap
PURPOSE:
Swapping bytes
CATEGORY:
gen/for/lib
CALLING SEQUENCE:
subroutine CvSwap(IB,NB,NN,X)
INPUTS:
IB integer # bytes treated as single unit (1,2 or 4)
NB integer # bytes in one element of X (1,2,4 or 8)
(1 for byte, 2 for integer*2, 4 for integer*4 and real*4,
8 for real*8)
NN integer # elements in X
X(NB,NN)integer*1 declared as integer*1 array, but can be anything
OUTPUTS:
X(NB,NN)integer*1 input array with bytes swapped
CALLS:
(none)
CALLED BY:
CvI2, CvI4, CvR4, CvR4S, CvR8, CvR8S, HOSInquire, OCv2Flt, iHOSRead, iHOSWrite
EXAMPLE:
For integer*4 array I(10)
call CvSwap(1,4,10,I)
will reverse the order of bytes in each element of I (bytes 1,2,3,4 -> 4,3,2,1)
call CvSwap(2,4,10,I)
will reverse the order of words in each element of I (bytes 1,2,3,4 -> 3,4,1,2)
PROCEDURE:
MODIFICATION HISTORY:
SEP-1998, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
[Previous]
[Next]
NAME:
cXCvarFormat
PURPOSE:
Provides format specification for Carrington variables.
CATEGORY:
I/O
CALLING SEQUENCE:
character*9 function cXCvarFormat(NCoff,XC)
INPUTS:
NCoff integer Carrington rotation offset
XC real Carrington variable with NCoff subtracted
OUTPUTS:
cXCvarFormat character*9 string of type 1234.5678 (i.e. F9.4 format)
CALLS: ***
XCvarFormat
CALLED BY:
BListAll, MapReadSingle, MapReadTimes, T3D_Read_B
MODIFICATION HISTORY:
OCT-2003, Paul Hick (UCSD/CASS; pphick@ucsd.edu)
