plwin.c

Decoding and calculating lag profiles. And more. Using fft. Data vector may be zero-padded on both lower edge and upper edge of range. Can work with fractional lags.


Layout of parameter file:

Line	Name			Notes
----------------------------------------------------
0	npar			Number of parameters in the parameter file
1	fft_len			Size of fft
2	nr_gates		Number of range gates
3	nr_undec 		Number of undecoded gates
4	nr_win 			Number of different subsets in the code (number of data windows)
5	nr_rep 			Total number of subcycles in one integration period
6	nr_samp 		Number of samples per subcycle
7 	win_len 		Number of bits in the code
8 	frac 			Number of samples per code bit
9 	res_mult 		Number of subsets of code (normally 1)
10 	nr_pp - nr_samp		Number of extra samples in the power profile
11	lower_tail / frac 	Number of padded zeros at lower edge, code bit units
12 	upper_tail / frac 	Number of padded zeros at upper edge, code bit units
13 	undec1 			Start gate for undecoded
14 	undec2 			Stop gate for undecoded
15 	debug 			If >0: Display debugging messages
				If >1: Display more debugging messages
16 	maxthreads 		Maximum number of threads in the calcuclations
17 	dl_short 		Number of short D-layer lags
18 	dl_long 		Number of long D-layer lags
19 	nr_dgates 		Number of D-layer gates (including lower tail)
20 	---			Offset in input data vector
21 	do_pp 			If non-zero: Calculate power profile
22 	nr_clutter 		Number of points to de-clutter
23 	notch 			Filter width for clutter reduction (actual width: 2*notch+1)
				If negative: Use value provided through EROS
>23 	---			The code


Total number of lines in par-file:
24+nr_win*win_len


Output format:

Short description			Size (number of elements)
------------------------------------------------------------
Normal short lags			[(nr_gates+1)*frac]
Normal long lags 			[nr_gates*fft_len/2]
Normal undecoded lags			[nr_undec*frac]
Normal power profile			[do_pp*nr_pp]
Pulse-to-pulse lags			[nr_dgates*dl_short]
Pulseset-to-pulseset lags		[nr_dgates*dl_long]
Pulse-to-pulse coherent profile	[nr_dgates]

	The first four are repeated if res_mult>1.


Ordering of lags:
[lag1 range1], [lag1 range2] , ..., [lag1 rangeMAX], [lag2 range1], [lag2 range2], ...