I'm working on modernizing Rosetta Code's infrastructure. Starting with communications. Please accept this time-limited open invite to RC's Slack.. --Michael Mol (talk) 20:59, 30 May 2020 (UTC)

Talk:Find the intersection of two lines

From Rosetta Code

a REXX version of a REXX version[edit]

This REXX version is a re-write of version 2 of the REXX entry,   with:

  •   added the required comment so that this REXX version would execute on VM/CMS and MVS/TSO
  •   aligned indentation for all   do-end   blocks   (and encapsulated statements)
  •   elided distracting superfluous zeros in integers
  •   elided superfluous decimal points in integers
  •   elides the superfluous and distracting use of concatenation   (││)
  •   a unique symbol instead of a null literal for a special case
  •   aligns the data points and results in the output
  •   adds whitespace to make arithmetic computations more perusable
  •   eschews title-case capitalizations
  •   maintains the same line for the   then   clause and the   if   clause   (no split statements)
  •   uses indentations for all REXX statements in the function
  •   has the result on the same line as the input   (data points)
  •   a different quoted literal style   (for easier reading of multiple literals on the same clause)
  •   REXX variables to hold long literals that would otherwise cause excessive wide REXX statements
  •   a comma   (instead of a slash)   to separate the   x   y   coördinates of the data points.
  •   elides superfluous   do-end   block structures;   less clutter, easier to read
  •   added more whitespace within some REXX statements and the REXX program's output
  •   tests all data possibilities   (for showing all the tested non-intersecting conditions)
  •   for viewing the calculation in its entirity, all program logic was kept within single viewable screen
Translation of: REXX (version 2)
/*REXX program finds (possibly) the intersection of two lines  (with diagnostic errors).*/
say iSect( 4 0 6 10 0 3 10 7 )
say iSect( 0 0 0 10 0 3 10 7 )
say iSect( 0 0 0 10 0 3 10 7 )
say iSect( 0 0 0 1 1 0 1 7 )
say iSect( 0 0 0 0 0 3 10 7 )
say iSect( 0 0 3 3 0 0 6 6 )
say iSect( 0 0 3 3 0 1 6 7 )
say iSect( 0 0 3 3 8 8 8 8 )
exit /* ═══a═══ ═══b═══ ═══c═══ ═══c═══ stick a fork in it, we're all done. */
iSect: procedure; parse arg xa ya xb yb xc yc xd yd /*optain args from invocation. */
$=. /*the intersection or error msg.*/
if xa=xb then do; k1=. /*the slope is infinite. */
x1=xa /*the X's intersection is XA */
if ya=yb then $= 'points A and B are identical'
end /* [↑] AB is a vertical line. */
else do; k1=(yb-ya) / (xb-xa) /*compute the slope of AB */
d1=ya - k1 * xa /*calc. intersection with Y axis*/
end /* [↑] AB isn't a vertical line*/
if xc=xd then do; k2=. /*the slope is infinite. */
x2=xc /*the C's intersection is XC */
if yc=yd then $= 'points C and D are identical'
end /* [↑] CD is a vertical line. */
else do; k2=(yd-yc) / (xd-xc) /*compute the slope of CD */
d2=yc - k2 * xc /*calc. intersection with Y axis*/
end /* [↑] CD isn't a vertical line*/
@ident= 'lines AB and CD are identical' /*literal to help shorten a line*/
@paral= 'lines AB and CD are parallel' /* " " " " " " */
/* [↓] no special case so far···*/
if $=. then if k1=. then if k2=. then if x1=x2 then [email protected] /*identical. */
else [email protected] /*parallel. */
else do; x=x1 /*use X1 */
y=k2 * x + d2 /*Y from CD */
else if k2=. then do; x=x2 /*X from CD */
y=k1 * x + d1 /*Y from AB */
else if k1=k2 then if d1=d2 then $= @ident
else $= @paral
else do; x=(d2-d1) / (k1-k2)
y=k1 * x + d1
end /* [↑] normal*/
if $=. then $= 'intersection is at (' || x","y')' /*$ ¬defined?*/
@ = left( 'a=('xa","ya')', 12) left( 'b=('xb","yb')', 12), /*whitespace.*/
left( 'c=('xc","yc')', 12) left( 'd=('xd","yd')', 12) /* " */
return left(@, max(51, length(@) ) ) ' ───► ' $ /*return str.*/
output   when using the default inputs:
a=(4,0)      b=(6,10)     c=(0,3)      d=(10,7)      ───►  intersection is at  (5,5)
a=(0,0)      b=(0,10)     c=(0,3)      d=(10,7)      ───►  intersection is at  (0,3)
a=(0,0)      b=(0,10)     c=(0,3)      d=(10,7)      ───►  intersection is at  (0,3)
a=(0,0)      b=(0,1)      c=(1,0)      d=(1,7)       ───►  lines  AB  and  CD  are  parallel
a=(0,0)      b=(0,0)      c=(0,3)      d=(10,7)      ───►  points  A  and  B   are identical
a=(0,0)      b=(3,3)      c=(0,0)      d=(6,6)       ───►  lines  AB  and  CD  are identical
a=(0,0)      b=(3,3)      c=(0,1)      d=(6,7)       ───►  lines  AB  and  CD  are  parallel
a=(0,0)      b=(3,3)      c=(8,8)      d=(8,8)       ───►  points  C  and  D   are identical
If I were to translate many of the Rexx programs of GS to my liking I'd be busy for months.
  •  Here is the Rexx program as I wrote it in my style:::*  


  •  Any opinions from any peers? --Walterpachl (talk) 06:01, 19 May 2017 (UTC)

There is no reason to take it personally.   It wasn't meant as a criticism, it is just another version in a different style, albeit a fair number of (style) differences.   Everybody's code can be improved (as least, the style can be changed).   In this case, I elided a few superfluous statements, which, in my opinion, didn't add anything to the REXX program or make it easier to understand/peruse.   I didn't appreciate your style of capitalization, misaligned DO-END statements (and the intervening/encapsulating REXX statements), split IF-THEN clauses, and much more.     But, that's only my opinion and preferences,   I merely added a version that I found easier to read and understand (and I hoped others will appreciate this version), and I also removed superfluous DO-END blocks and such.   Note that this re-written REXX version was added in the   discussion   section, not on the   page   section so as to not clutter up the main page.   I found that that particular REXX version was so difficult to follow and understand (the IF logic) with all the multiple misaligned DO-END and compound IF THEN-ELSE statements.   The version (above) that I re-wrote speaks to style and understandability.   There are many styles to write REXX programs in, and this is just one of them.   Nobody's style is everybody's cup of tea.   -- Gerard Schildberger (talk) 08:22, 19 May 2017 (UTC)

The only thing I liked about this variation is the presentation of the result.
The a= etc. should be A= etc. to be consistent.
I dislike that the variation can no longer be used with ooRexx :-(
and i cannot appreciate the landscape formatting!
But let's agree that our taste as far as formatting is concerned is vastly different.
And I added commentary to my version. --Walterpachl (talk) 09:29, 19 May 2017 (UTC)

Clojure version does not handle edge case case of undefined slope (vertical line)[edit]

The part that calculates m (slope) fails if `(- x2 x1)` is zero.

 (defn compute-line [pt1 pt2]
   (let [[x1 y1] pt1
         [x2 y2] pt2
         m (/ (- y2 y1) (- x2 x1))]
     {:slope  m
      :offset (- y1 (* m x1))}))

Should the clojure version do something other than throw an exception if the compute-line fn receives a vertical line like [0 0] [0 6], but also all cases like [X Y] [X Z], I would suspect.

Currently it just throws on divide by zero, which means you can not find the intersection if the both x-coords points of one line are the same.


can we have a webpage version of the code?

I just added an online link for a transpiled version of Phix (am currently doing that sort of thing to lots of pages), not exactly what you asked for I know, but you could use that as a starting point: rip out my p2js.js, make X=0 and Y=1, replace $subse() with plain e[X] etc, kill off all those ugly "sequence", a few "[," ==> "[", and lastly you'll need a replacement for my print(). HTH. --Pete Lomax (talk) 21:04, 29 December 2021 (UTC) PS Please in future sign your posts using --~~~~ (without the nowiki bits).