Go to the previous, next section.
Many commands exist in UniPOPS to assist you with operations you may require in the reduction of your observational data. Most of these will be described below. However, you should refer to Chapter 7 for a description of the UniPOPS facilities for fitting Gaussians and baselines, and Chapter 8 for a description of the smoothing of data, plus the verbs for convolution and cross-correlation,
While you should have some information on your observations from the telescope log sheets, and (we trust) from your own observing notes, UniPOPS provides a number of facilities to augment this, aid the human memory and even update the header parameters of a scan, or its data. A number of verbs print standard information about the scan on the screen, while others allow you to check the more esoteric parameters associated with these data. You should be aware that a number of these verbs behave differently depending upon whether the data was taken at Tucson or Green Bank.
Tucson Version : Scan number, LST, source name, rest frequency, filter-bank and sideband number.
Green Bank Version : Scan number, LST, source name, rest frequency and channel separation (in MHz).
To invoke this verb, type,
>TITLE
To make a table for many scans with DOC, either a FOR-loop (see Section 13.5), or the stack (see Chapter 11) are useful. For example, to list parameters for scans 1000 through 1020, type,
>SCALAR SCAN_I >PFLAG = FALSE >FOR SCAN_I = 1000 TO 1020; GET SCAN_I; DOC; END
If, instead, you want to produce this list for scans 602, 604, 606, and 1000 through 1020, it is easiest to use the stack by typing,
>SCALAR SCAN_I
>EMPTY
>ADDSTACK(602, 606, 2) >ADDSTACK(1000, 1020, 1) >PFLAG = FALSE >FOR SCAN_I = 1 TO ACOUNT; GET ASTACK(SCAN_I); DOC; END
DOC will list scan number, observing date (YY.MMDD), LST, RA, Dec, scan duration (sec), system temperature (Tsys), rest frequency and source name.
>HEADER
Tucson Version : THUMB prints the main and reference azimuth, elevation, RA and Dec offsets.
Green Bank Version : THUMB prints two lines describing the pointing corrections. The first line contains the PVLS (the box installation corrections : P1, P2 and P3), while the second line gives the LPC's (the Local Pointing Corrections : RLPC and DLPC).
To invoke THUMB, type,
>THUMB
Currently, approximately one hundred and forty header parameters can be interrogated, and even updated, using the facilities of UniPOPS. This is achieved via keyword adverbs (listed via EXPLAIN SYNOPSIS and in APPENDIX A.4). The keyword adverbs are atypical and cannot be changed by assignment statements. They serve purely as pointers to scan header elements, and are used as arguments for the internal-array verbs H0 - H9. Some header parameters are normal reals, others are strings, while yet others are arrays of reals or strings. Some keyword adverbs apply only for Green Bank headers, some only for Tucson headers, and some for both. The use of the verbs H0 - H9 for the various keyword adverbs is as follows,
i) To interrogate the header using the various keyword adverbs, use the verb PRINT (see Chapter 14.5), or its alias ?, with the appropriate calls to the internal-array verbs Hn as attributes. The internal-array verb Hn, where n is the number of the UniPOPS Array in which the scan is held, takes the appropriate keyword adverb as argument. For example, if there is a scan in Array (4) and you wish to find what the ambient temperature (keyword adverb TAMB) was at the time of the observation, type,
>PRINT H4(TAMB)
or
>? H4(TAMB)
To set an updated value in a header parameter, again use the verbs Hn, with the appropriate keyword adverb as argument, and make it the left-hand side of an assignment statement, with the new value as the right hand side. For example, if you want to update the value of the ambient temperature to 5 deg C in the header of the scan of the above example, then type,
>H4(TAMB) = 5.
Further examples might be,
>SCALAR GAL_LAT, T_AMBIENT >GAL_LAT = H0(GALLAT) >T_AMBIENT = 28 >H4(TAMB) = T_AMBIENT >H8(TAMB) = H0(TAMB)
ii) To interrogate a header parameter which is an array of reals, again use the internal-array verb Hn, where n is the number of the relevant UniPOPS Array, with the appropriate keyword adverb plus an increment of (m - 1) as argument, to print the m-th array element. For example, should you wish to find the center of the descriptive coordinate system (keyword adverb DESORG pointing at a 3-element real array) in which the observation was made for a scan in Array (6), type,
>PRINT H6(DESORG), H6(DESORG + 1), H6(DESORG + 2)
Similarly, to reset the third element of this array to 32 deg for the header of the scan in Array (6), type,
>H6(DESORG + 2) = 32.
iii) To interrogate a simple string in a header, using the keyword adverb which points to the relevant string, again use the internal-array verb Hn, with the appropriate keyword adverb as argument. Suppose you wish to find the observer's name (keyword adverb OBSERVER) from the header of the scan in Array (0), type,
>PRINT H0(OBSERVER)
To reset a string in the header of a scan, use the verb Hn, with the keyword adverb pointing at the string as argument, as the left-hand side of an assignment statement, and the replacement value as the right-hand side. For example, if you wish to reset the observer's name in the header of the scan held in Array (0), type,
>H0(OBSERVER) = "Bloggs"
Note that in this statement, Hn can take a character literal, a string adverb, or a string function verb as its second argument. For example,
>YTITLE = H0(BACKEND) >XTITLE = 'GBT'; H0(TELESCOP) = XTITLE >H4(TELESCOP) = H0(TELESCOP)
You should be aware that it is possible to change the header values so that they are no longer consistent with other header values. For example, the frequency information (observed frequency, OBSFREQ, rest frequency, FREQRES, and frequency resolution, FREQRES) are related to the the velocity information (the velocity at the reference pixel, VELOCITY, the Doppler correction for the earth's motion, RVSYS, and the velocity definition, VELDEF) as well as the x-axis descriptive information (the reference channel, REFPT, the value at the reference channel, X0, and the increment between channels, DELTAX). UniPOPS always attempts to derive velocities from the frequency information and the value of the reference channel, REFPT plus the Doppler correction, RVSYS. It is therefor possible to affect the displayed and calculated values for velocity without actually changing the values for VELOCITY, X0, and DELTAX stored in the header. While this may not be an immediate problem, months or years from now when you attempt to interpret this data, possibly without the help of UniPOPS, you may not know which information to believe. The responsibility for keeping the information in the header self-consistent is entirely the user's. UniPOPS will allow you to change any header value independent of any other header value.
The data values of all, or part, of the scan held in Array (0) are listed simply by using the verb TABLE. TABLE will list a header giving the scan number and channel separation. This is followed by the actual table of data, giving x-axis values (in units of the lower x-axis) and antenna temperature for each point. These are printed with three entries per line. The section of the data array to be listed can be controlled by the adverbs BDROP and EDROP, which specify the number of points to be ignored at each end of the scan.
For example, suppose the scan in Array (0) contains 512 data elements, to list all of these, type,
>BDROP = 0 ; EDROP = 0 >TABLE
If it were required to list only elements 100 to 200, type,
>BDROP = 99; EDROP = 312
>TABLE
More flexibly, the internal-array verbs D0 - D9 can be used, with the verb PRINT, to list the data in a format of the user's choice. The verb Dn(m) will return the data value in the m-th channel in Array (n). Thus, to print elements 100 to 200 of the spectrum held in Array (2), in a "four-to-a-line" format, type,
>FOR I = 100 TO 195 BY 4; ? D2(I) D2(I+1) D2(I+2) D2(I+3);\
END; ? D2(200)
(Note the use of the back-slash to continue the command line.)
If you want to modify, or "paint in", data values, for example to replace interference spikes using the algorithm of your choice, or to create a "dummy" scan, then the internal-array verbs D0 - D9 can again be used. The relevant Dn(m) for the m-th point of the scan held in Array (n) should be placed as the left-hand side of an assignment statement, with the required new value as the right-hand side. Suppose you wish to set an interfered 100-th channel of the spectrum in Array (1) to the mean of the adjacent channels, type,
>D1(100) = 0.5 * (D1(101) + D1(99))
Alternatively, if you wish to modify the data in Array (0) by dividing each value by the square root of the equivalent point in Array (2), type,
>SCALAR P_I >FOR P_I = 1 TO H0(NOINT); D0(P_I) = D0(P_I)/SQRT(D2(P_I)); END > # or {D0@ = D0@/SQRT(D2@)} (see Section 14.4)
You can also give any element of a data array an undefined value. Most verbs will know how to ignore or use a data point that is undefined. An undefined data value is one that is equal to IEEE Inf (infinity). You can set a data value to be undefined by setting it equal to the special adverb, DEFAULT. For example, to give channel 42 an undefined value, you type:
> D0(42) = DEFAULT
The verb SCALE allows you to multiply the data in Array (0) by a constant factor, while the verb BIAS will add a constant value to each data point in Array (0). These verbs share a common adverb, FACT.
The verb SCALE multiplies the data in Array (0) by the numerical value held in the adverb FACT. Suppose you wish to multiply the current scan in Array (0) by 3.2, type,
>FACT = 3.2 >SCALE
The verb BIAS adds the value held in the adverb FACT to each element of data held in Array (0). Thus, if you want to display scans 1600 and 1610 on the same plot, with the second of these displaced vertically for clarity, you might type,
>GET 1600 ; PAGE SHOW >GET 1610 >FACT = 3. >BIAS RESHOW
Sometimes, you may wish to perform simple math manipulations between two scans; for example, the necessary differencing of ON and OFF total-power scans. To achieve this, we have provided five verbs PLUS, MINUS, DIFF, MULTIPLY, and DIVIDE. The verb DIFF and MINUS are identical (for compatibility with older versions of the program) will subtract the scan in Array (1) from that in Array (0), placing the result in Array (0). Suppose that scans 700 and 702 are an ON-OFF pair. To difference them, type,
either
>GET1 702 ; GET 700 >DIFF
or
>OFF 702; ON 700 >DIFF
or
>GET 702 ; COPY(0,1) GET 700 >DIFF
All three of these solutions will have the same effect.
To divide the data in Array (0) by those in Array (1), placing the result in Array (0), the verb DIVIDE should be used. Suppose it is wished to calibrate (in K) the differenced data obtained in the last example (see above) by dividing this differenced data by the OFF scan (already present in Array (1)), and multiplying by the system temperature. Then continue the above example by typing,
>DIVIDE >FACT = H1(STSYS) ; SCALE
The verbs MULTIPLY and PLUS act in an obvious way on the contents of Arrays (0) and (1). Again, the results overwrite the contents of Array (0).
All of these array math verbs use the value of the DEFMODE adverb to determine how they treat undefined data values. An undefined data value is one that is equal to IEEE Inf (infinity).
If DEFMODE is FALSE, the result of an array math verb will be undefined (IEEE Inf) at any location where either of the values in Array (0) or Array(1) are undefined.
If DEFMODE is TRUE, the result of an array math verb will be undefined at any location where both of the values in Array (0) or Array(1) are undefined and it will use the defined value only in channels where only one of the values in Array (0) or Array (1) are undefined. See the documentation on the specific verb for details about the behavior when DEFMODE is TRUE.
The default value for DEFMODE is FALSE. This is likely to be the behavior that most users want.
In practice, a special verb TEMP exists especially for producing temperature-normalized differenced data in Array (0) for an ON-OFF pair. TEMP performs slightly different operations depending on whether the observations are from Green Bank or Tucson.
Green Bank Version : TEMP computes,
((ON - OFF) / OFF) * TSYS
Where, ON = Total-power ON scan held in Array (0), OFF = Total-power OFF scan in Array (1), TSYS = System temperature of the ON scan.
This is the same operation as was performed by the combination of the previous two examples. Using TEMP to achieve this operation, type,
>GET1 702 ; GET 700 >TEMP
Tucson Version : TEMP computes,
((ON - OFF) / OFF) * GAINS
Where, ON = ON scan in Array (0), OFF = OFF scan in Array (1), GAINS = GAINS array in Array (2).
The scan held in Array (0) can be reversed, end-to-end, using the verb INVERT. Of course, a second application of INVERT will return Array (0) to its original state. Suppose it is required to flip scan 700 end-to-end and display it, type,
>GET 700 >INVERT PAGE SHOW
The verb FFT performs a fast Fourier Transform (FFT) of the data held in Array (0), placing the real part of the result in Array (0), and the imaginary part in Array (1).
The inverse Fourier Transform of a complex data set stored with its real part in Array (0), and its imaginary part in Array (1), is performed by invoking the verb IFFT. IFFT places the real part of the result in Array (0), and the imaginary part in Array (1).
Both FFT and IFFT use the adverbs BDROP and EDROP, ignoring these numbers of channels at each end of the spectrum.
Suppose that you want to take the Fourier Transform of scan 700, display both real and imaginary parts of the result, and then take the inverse transform to restore the "status quo". Type,
>GET 700; PAGE SHOW >FFT PAGE SHOW >MOVE(0,2) COPY(1,0) PAGE SHOW >MOVE(2,0) IFFT >PAGE SHOW
Once the Fourier Transform or the inverse Fourier Transform has been obtained via FFT or IFFT, it is possible to compute the power in each channel, (the power spectrum of the original data), using the verb POWSPEC. POWSPEC takes the sum of the the squares of the real part in Array (0) and the imaginary part in Array (1) to give the power spectrum, placing the result in Array (0). POWSPEC also ignores BDROP and EDROP channels at each end of the spectrum.
You may also want to indicate to POWSPEC how undefined data values are to be treated by setting the DEFMODE adverb to be either FALSE (the default value) or (TRUE). An undefined data value is a one which is equal to the IEEE Inf (infinity) value. If DEFMODE is FALSE, POWSPEC will place an undefined value at any channel where either of the values in the two input arrays are undefined. If DEFMODE is TRUE, POWSPEC will put an undefined value at any channel where both of the values in the two input arrays are undefined and will place the valid value in a channel if only one of the two input values at that channel are valid. Most users will leave DEFMODE at its default setting of FALSE.
For example, to display the power spectrum of scan 700, then type,
>GET 700 ; BDROP = 0 ; EDROP = 0 >FFT POWSPEC >PAGE SHOW
At all telescopes, scans are affected by interference, instrumental effects, etc. It is usually required to minimize these effects, and possibly make an "inspired guess" as to what the corrupted data should have contained. Three somewhat-different approaches to this problem are available in UniPOPS. In addition, you can always "paint in" data values, as described in Chapter 9.5.
The simplest remedy for spikes in the data is to use the verb CLIP. CLIP locates all data values in Array (0) that are above or below the contents of the adverbs CLIPMAX and CLIPMIN respectively, and replaces them with CLIPMAX or CLIPMIN, as appropriate. CLIP makes no attempt to guess the likely value of the corrupted point, but its use will permit the automatic scaling facility in SHOW to do full justice to the data. CLIP can also be used to truncate an intense spectral feature, allowing automatic scaling in SHOW to highlight weaker lines or low-level structure in the spectrum.
Suppose that a large interference spike (say 100 K in height) dominates the line under observation (2 K). Then, to produce a reasonable display from SHOW to bring out the astronomical content of the spectrum, type,
>CLIPMIN = -1. ; CLIPMAX = 2.5 >CLIP PAGE SHOW
A more sophisticated treatment of corrupted data can be achieved using the combination of the verbs SPIKE and REPLACE. SPIKE searches Array (0) for data whose ABSOLUTE intensity values are greater than the adverb CUTOFF. Up to 12 such channels can be accommodated and their channel numbers will be both printed in the text window, and set into the array adverb BADPT, for use with the verb REPLACE (see below). Note that this verb leaves zeros in all elements of BADPT not required for "bad" channels. The adverbs BDROP and EDROP define the number of channels that SPIKE ignores at the beginning and end of the scan.
The verb REPLACE should be invoked when you are sure that BADPT contains only the channels that you want replaced. (If you do not like what SPIKE has done, you can always set, or modify, the array BADPT yourself using, for example, a BADPT = .... , or BADPT(n) = .... assignment statement.) REPLACE replaces the data of the channels held in BADPT by "best guess" values derived from parabolic interpolation over adjacent channels. The adverbs BDROP and EDROP define the number of channels that are ignored at the beginning and end of the scan, even if an entry in BADPT falls within one of those regions.
Suppose that you are considering the same scan with the 100 K spike, and the 2 K spectral line. To replace the spike with a "best-guess" value, type,
>BDROP = 0 ; EDROP = 0 >CUTOFF = 2.5 ; SPIKE >REPLACE PAGE SHOW
As a further example, suppose you found that channel 25 of your spectra was always bad. You could easily replace this on regular basis by first typing,
>BADPT = DEFAULT >BADPT(1) = 25
and then after GETting each spectrum, typing,
>REPLACE
Finally, one can simply choose to make the data undefined at that channel by setting it equal to the special adverb, DEFAULT. For array data, an undefined value is any value equal to IEEE Inf (infinity). Verbs are able to deal sensibly with such data (generally by ignoring it) although some verbs will insist that all data values be defined (valid). The behavior of the array math verbs (Section 9.7), POWSPEC (9.9), RAP (9.12), SUM and ACCUM (Chapter 10) as well as the matrix math verbs (16.6.2.3) with respect to undefined values can be altered by setting the adverb DEFMODE. For example, if the value in channel 72 is bad, and you want to set this value to be undefined, you type:
>D0(72) = DEFAULT
The verb MOMENT computes the zeroth and first moment for chosen regions of the scan in Array (0), storing the result in the two-element array adverb SIZE. The adverbs BMOMENT and EMOMENT contain the first and last channels of the region of interest. After MOMENT has been invoked, SIZE contains,
SIZE(1) = zeroth moment, i.e. the area under the curve in units of the y-axis times that of the lower x-axis. SIZE(2) = first moment, i.e. the intensity-weighted x-value in units of the lower x-axis.
(REMEMBER : For the results of MOMENT to be meaningful, the spectrum must have had its baseline optimized before invoking the verb.)
Suppose you wish to obtain the first two moments for the narrow spectral line at channel 135 for the already-baselined spectrum in Array (0), and to draw a vertical line on the plot at the position of the line center. Then type,
>BMOMENT = 125 ; EMOMENT = 145 >SAXIS(VELHEAD,CHANNEL) MOMENT >PRINT SIZE >FLAG(SIZE(2))
(Acknowledgement : The code and documentation for GMEASURE, described here, are adapted from the Galpac package in the ANALYZ program.)
The verb GMEASURE allows you to compute a flux integral (in units of the y-axis times that of the lower x-axis), a center and a width (both in units of the lower x-axis) for the spectrum of a spiral galaxy in Array (0). GMEASURE takes a single argument, consisting of an integer of up to five digits which act as codes for the types of operation you require GMEASURE to perform (see below). The 5-element array adverb GMFRAC contains input values as detailed below, while the 6 x 5-element array adverb GOUT holds the results.
Before using GMEASURE, make sure that,
1) The scan you are applying it to has had a baseline removed.
2) You have used the verb RMS to get a value of the rms of the data off the spectral line. The rms value is placed into the adverb VRMS (see Section 7.6).
3) You have set the adverbs BMOMENT and EMOMENT to the first and last channels of the region of interest. Alternatively, if you set,
BMOMENT = EMOMENT = a channel within the spectral feature,
then the program will search for, and use, the channel numbers of the first negative data values to the left and right of the channel number you supplied in BMOMENT.
GMEASURE has five different operation modes for computing the flux integral, velocity, and width, as follows,
| Code | Operation Mode |-------|--------------------------------------------------------- | 1 | At a fraction f of the MEAN flux over the region of | | interest defined by BMOMENT and EMOMENT. |-------|--------------------------------------------------------- | 2 | At a fraction f of the PEAK flux over the region of | | interest defined by BMOMENT and EMOMENT. |-------|--------------------------------------------------------- | 3 | At a fraction f of each of the two "HORNS", searched | | from BMOMENT and EMOMENT towards the center of the | | feature. |-------|--------------------------------------------------------- | 4 | At a fraction f of each of the two "HORNS", searched | | from BMOMENT and EMOMENT in, and from the center of the | | feature out and the average of the two reported. A | | polynomial is fitted between fractions f1 and f2 (see | | below) of the two horns, and the coefficients used to | | determine the velocity uncertainty. |-------|--------------------------------------------------------- | 5 | At a fraction f of each of the two "HORNS", searched | | using an interpolating polynomial between fractions f1 | | and f2 (see below) of each horn. The velocity | | uncertainty is estimated from the covariance matrix of | | the fitted polynomial coefficients.
The argument you supply to GMEASURE permits it to make up to 5 simultaneous estimates of the velocity and width. The flux integral will be the same for each estimate (since it is the integral between BMOMENT and EMOMENT). The digits of the argument which you supply identify the code number for the required mode of operation (from the above table) for each estimate. Associated with each digit in the argument, you should set the values of the corresponding element of array GMFRAC, as follows,
a) If the i-th digit of the argument = 1 to 3, GMFRAC(i) = f. b) If the i-th digit of the argument = 4 or 5, GMFRAC(i) contains a value given by,
GMFRAC(i) = f1 + 0.1 * f2 + 0.01 * f
For example, if you invoked GMEASURE as follows,
>GMFRAC = 0.5, 0.2, 0.5, 0.825, 0.915 >GMEASURE(12345)
GMEASURE would display the flux integrals, centers and widths, computed for,
i) 50 % of the mean (Mode 1, GMFRAC(1)).
ii) 20 % of the peak (Mode 2, GMFRAC(2)).
iii) 50 % of the two peaks (Mode 3, GMFRAC(3)).
iv) 50 % of the two peaks, with the error estimated from a
polynomial fit between 20 and 80 % of the two peaks.
(Mode 4, GMFRAC(4) = 80 % + 0.1 * 20 % +0.01 * 50 %).
v) 50 % of the two peaks searched by fitting a polynomial
between 10 and 90 % of the two peaks, with the error
estimated from the fitted polynomial.
(Mode 5, GMFRAC(5) = 90 % + 0.1 * 10 % + 0.01 * 50 %).
Note : The difference between the modes 4 and 5 is that mode 4 locates the 50 % point by linear interpolation between channels bracketing the 50 % height, while mode 5 uses the polynomial to locate the 50 % point. Mode 5 is intended for use on high S/N-ratio data.
For modes 4 and 5, the following algorithm determines the peaks in the profile between BMOMENT and EMOMENT.
a) All peaks in this range are identified. A peak is defined as a channel with more flux than a number of channels selected by GMEASURE (between 3 and 7) on either side of it.
b) For only one peak, the profile is treated as double-horned
with coinciding horns.
c) For only two peaks, the profile is treated as double-horned.
d) For more than two peaks, the two outermost peaks are treated as the horns UNLESS,
e) One of the inner peaks has more than twice the flux of the nearest outer peak. That inner peak is then treated as the horn. If more than one inner peak meets this criterion, then the outermost is taken as the horn.
The array adverb GOUT, which contains the results for the (up to) five estimates (i = 1 to 5) in its 30 currently-used elements, is organized as follows,
GOUT(i, 1) = Mode number for i-th estimate. GOUT(i, 2) = center for the i-th estimate. GOUT(i, 3) = Width for the i-th estimate. GOUT(i, 4) = center error (modes 4 & 5 only) for the i-th estimate. GOUT(i, 5) = The value in GMFRAC(i) for the i-th estimate. GOUT(6, 1) = Peak flux value within region of interest. GOUT(6, 2) = The flux integral. GOUT(6, 3) = RMS value as found initially in adverb VRMS. GOUT(6, 4) = BMOMENT GOUT(6, 5) = EMOMENT
Note : The reported center error is the estimated error in the average center of the profile, the error in the width is sqrt(2) larger.
If you wish to combine the left half of a frequency-switched scan in Array (0) with the right half, then use the verb RAP. On completion, RAP places the result in Array (0), and sets the adverb OSHIFT equal to the number of channels between Signal and Reference frequencies. (At present, RAP only works correctly if the frequency shift is an integral number of channels.)
Be careful when using RAP, as its operation is rather different for Green Bank and Tucson data.
Green Bank Data : RAP will compute the number of channels to
"shift and combine" from the frequencies stored in the scan header. Thus to perform the combination, just type,
>RAP
Tucson Data : Here, one must set the two adverbs, FS (Frequency
offset in MHz for the Signal data) and FR (Frequency offset in MHz for the Reference data) before invoking RAP. As of Fall 1993, these two values are available from the header. There are three frequency offsets stored in the header: FOFFSIG, FOFFREF1, and FOFFREF2. Usually you will want to set FS to the value of the FOFFSIG header word and FR to the value of the FOFFREF1 header word.
> FS = H0(FOFFSIG) > FR = H0(FOFFREF1) > RAP
NOTE : Following the application of RAP, the user must decide which half of the spectrum is the signal half, and drop the other half via the adverbs BDROP and EDROP before displaying the data. If OSHIFT is > 0, then you should set BDROP to at least the value of OSHIFT. If OSHIFT is < 0, then you should set EDROP to at least the absolute value of OSHIFT.
The adverb DEFMODE indicates how RAP treats undefined data values. An undefined data value is one equal to IEEE Inf (Infinity).
If DEFMODE is FALSE, RAP will place an undefined data value at any location where either the signal or reference values are undefined (i.e. if channel 100 is undefined, as above, then the channel in the result that depends on channel 100 will also be undefined if DEFMODE is FALSE).
If DEFMODE is TRUE, RAP will place an undefined data value at any location where both the signal and reference values are undefined. However, if only one value is undefined, then the result will be as if the undefined value were zero (i.e. if channel 100 is the only undefined value, then that value will be treated as if it were 0).
Most users will leave DEFMODE set to it's default value of FALSE.
Two UniPOPS function verbs, CHANTOX and XTOCHAN, allow you to convert from the units of your lower x-axis into channel number, or vice versa. The units of the quantity that will be converted to or from channel number is defined by the most recently-set units for the lower x-axis, decided by a call to any of the verbs SAXIS, CC, CF, CV, FC, FF, FV, VC, VF or VV (see Section 6.2). CHANTOX takes a single argument representing a channel number, while XTOCHAN's single argument represents the value to convert into channel number. For example, if you wish to know which channel is at a radial velocity of +34.5 km/sec, and which channel is at a frequency of +1.25 MHz relative to the center of the bandpass, type,
>SAXIS(VELHEAD,CHANNEL); PRINT XTOCHAN(34.5) >SAXIS(FREQOFF,CHANNEL); PRINT XTOCHAN(1.25)
and the channel numbers will be printed. Alternatively, if you wish to know the velocity and frequency of channel 123, type,
>SAXIS(VELHEAD,CHANNEL); PRINT CHANTOX(123) >SAXIS(FREQOFF,CHANNEL); PRINT CHANTOX(123)
and again the answers will be printed.
Two UniPOPS verbs allow you to print the results of equatorial-galactic coordinate conversion on the text screen. To convert equatorial coordinates (epoch 1950) into galactic coordinates, the verb EQTOGAL is invoked. This takes two arguments, these being the RA(1950), Dec(1950) of interest, entered as shown below.
Suppose that you wish to convert the equatorial position of the peculiar star SS433 (RA50 = 19 09 21.3, Dec50 = +04 53 53.1) into galactic coordinates, then type,
>EQTOGAL(190921.3, 45353.1)
To convert a position in galactic coordinates into equatorial coordinates (epoch 1950), use the verb GALTOEQ, which takes the galactic coordinates (l,b), entered as shown below, as its two arguments.
Suppose that you wish to see the equatorial position of SS433 (l = 39.694, b = -2.245), type,
>GALTOEQ(39.694, -2.245)
Go to the previous, next section.
Webmaster: Ronald J. Maddalena