Calculate distances of continuous point - awk

Question

I need a help!
I have several points ABCDEF... with positions like this:
A 0.00 0.50 0.10
B 1.00 2.50 2.00
C 0.70 0.88 1.29
D 2.13 2.90 0.11
E 1.99 0.77 0.69
...
I aim to calculate distance of AB, BC, CD, EF,... and sum of them with output which has form like this:
sum_distance(AB)
sum_distance(AB+BC)
sum_distance(AB+BC+CD)
sum_distance(AB+BC+CD+DE)
sum_distance(AB+BC+CD+DE+EF)
....
I found on internet awk can do it and apply to my case. However, no result or error was exported on screen. Could you please help me with this situation?
bash shell, awk
awk 'FNR==NR { a[NR]=$0; next } { for (i=FNR+1;i<=NR-1;i++) {split(a[i],b); print $1 "-" b[1], sqrt(($2-b[2])^2 + ($3-b[3])^2 + ($4-b[4])^2) | "column -t" } NR--}'
Output:
2.934280150
4.728297987
7.470140434
9.682130488
11.92469598
......

Answer 1

You don't need such a complex script for this trivial task. Try this instead:
awk 'NR>1{ printf "%.9f\n",s+=sqrt(($2-x)^2+($3-y)^2+($4-z)^2) }
{ x=$2;y=$3;z=$4 }' file
For all points but A, calculate the distance, add it to the sum s and print the sum. For all points keep coordinates in x, y, z for next calculation. Its output looks like this with gawk:
2.934280150
4.728297987
7.470140434
9.682130488

Answer 2

What is the Cardinal Rule? (never use code off the internet you don't understand...)
The problem with the awk script you are attempting to use is it is not exactly your case. By setting FNR==NR and then using the loop limits (i=FNR+1;i<=NR-1;i++) it is expecting multiple input files. For your case, you can actually simplify the script by removing the loop entirely since you only have a single input file.
You need only save the first row, then using next read the next row, compute and output the distance between the prior row and the current, set the current row as the row in the a[] array and repeat until you run out of rows, e.g.
awk '{
a[NR]=$0
if (NR == 1)
next
split(a[NR-1],b)
printf "%s\t%s\n", b[1] "-" $1,
sqrt(($2-b[2])^2 + ($3-b[3])^2 + ($4-b[4])^2)
a[NR]=$0
}'
Example Input File
$ cat f
A 0.00 0.50 0.10
B 1.00 2.50 2.00
C 0.70 0.88 1.29
D 2.13 2.90 0.11
E 1.99 0.77 0.69
Example Use/Output
Simply paste the script into a terminal adding the filename at the end, e.g.
$ awk '{
> a[NR]=$0
> if (NR == 1)
> next
> split(a[NR-1],b)
> printf "%s\t%s\n", b[1] "-" $1,
> sqrt(($2-b[2])^2 + ($3-b[3])^2 + ($4-b[4])^2)
> a[NR]=$0
> }' f
A-B 2.93428
B-C 1.79402
C-D 2.74184
D-E 2.21199
Look things over and let me know if you have further questions.

Answer 3

Try this:
awk 'function d(a,b){split(a,x);split(b,y);return sqrt((x[2]-y[2])^2 + (x[3]-y[3])^2 + (x[4]-y[4])^2);} {p[FNR]=$0} FNR>1{sum[FNR]=sum[FNR-1]+d(p[FNR-1],p[FNR]);printf "%.9f\n",sum[FNR];}' file
With file content like this:
A 0.00 0.50 0.10
B 1.00 2.50 2.00
C 0.70 0.88 1.29
D 2.13 2.90 0.11
E 1.99 0.77 0.69
will provide output like this:
2.934280150
4.728297987
7.470140434
9.682130488
You didn't provide point F so your last line of output actually can't be count here.
Put into several lines here:
awk '
function d(a,b){
split(a,x);
split(b,y);
return sqrt((x[2]-y[2])^2 + (x[3]-y[3])^2 + (x[4]-y[4])^2);
}
{p[FNR]=$0}
FNR>1{
sum[FNR]=sum[FNR-1]+d(p[FNR-1],p[FNR]);
printf "%.9f\n",sum[FNR];
}' file
It's quite straightforward here, function d for distance. And reuse the sum of former line.
And for fun, if you want to calculate the total distance of a graph, with initially one point and gradually add point into the graph. I.E. :
sum_distance(AB)
sum_distance(AB+BC+AC)
sum_distance(AB+BC+AC+AD+BD+CD)
...
Then just a little bit improvement will do, like this:
$ awk 'function d(a,b){split(a,x);split(b,y);return sqrt((x[2]-y[2])^2 + (x[3]-y[3])^2 + (x[4]-y[4])^2);} {p[FNR]=$0} FNR>1{sum[FNR]=sum[FNR-1];for(i=FNR-1;i>0;i--)sum[FNR]+=d(p[i],p[FNR]);printf "%.9f\n",sum[FNR];}' file
2.934280150
6.160254691
14.349070561
22.466306583