format format format for t creative solution for Three Points Circle by veky - python coding challenges

Enable Javascript in your browser and then refresh this page, for a much enhanced experience.

format format format for t solution in Creative category for Three Points Circle by veky

def checkio(D):
    p=a,b,c=eval(D.replace("(","complex("))
    q=h,i,g=[p[i-2].conjugate()*(p[i-1]-p[i])for i in range(3)]
    s=(a*g+b*h+c*i)/sum(q)
    return"(x-{})^2+(y-{})^2={}^2".format(*((format(t,".{}g".format(3-(t<1)))
                                    for t in(s.real,s.imag,abs(s-a)))))

June 16, 2013

Comments:

bryukh on June 16, 2013, 7:59 p.m.
You solution does not work for x0 > 10, because "{:+.3g}".format(12.12) == "+12.1"

veky on June 16, 2013, 8:03 p.m.
I know. But I just wrote this intending to see if all numbers are one-digit-before-point, and it worked. In my view, it's a fault of test-writer. :-)

veky on June 16, 2013, 8:07 p.m.
Let me clarify: if you know a simple way to achieve required formatting, that also works with numbers >=10, I'll be glad to hear it. But I don't want to do microsurgery on my strings when obviously there are many things that can go wrong (-0, too big numbers in exponential format, 3.445 vs 3.455 rounding, etc.).

bryukh on June 16, 2013, 8:34 p.m.
I agree. I didnt find simple way for rounding without surgery.

AHrEJl on Aug. 13, 2013, 12:04 p.m.
Why not using round(...,2)? and %g output? worked pretty well in my solution

veky on Aug. 14, 2013, 9:52 a.m.
As I said, then too big numbers will still be written in exponential format. Mostly it just isn't worth it. :-)

bryukh on June 16, 2013, 8:34 p.m.
Yes, yes, yes :) My fault.

shreekar on Dec. 26, 2013, 8:09 a.m.
Neat solution. I completely forgot the complex number theory from high school and could not remember that it can be used to solve this problem.

veky on Dec. 26, 2013, 9:35 a.m.
Interesting. In my experience, much more frequent phenomenon is that people forget (or never learn actually) that Python has native support for complex numbers. :-)

ignalion on Feb. 10, 2014, 3:56 p.m.
This one won't work for x/y/r > 10. I think we can't expect '0-9 numbers only' just because it isn't stated otherwise.

veky on Feb. 10, 2014, 11:26 p.m.
See below. Requiring precise format without specifying ranges is bad spec. I refuse to follow it, especially if it means lenghtening my code unnecessarily - even if it costs me thumbs. :-]

bryukh on Feb. 11, 2014, 12:23 a.m.
Yes, i sprinkle ashes upon head :) I added the precondition about it and will add the new tests, when "tasks levels" is ready.

xxxxxx on July 19, 2014, 12:12 a.m.
Very nice. I don't get the idea behind the formula, though.

veky on July 19, 2014, 8:51 a.m.
It's a standard formula. You just solve |z-z1|=|z-z2|=|z-z3| (square, expand, rearrange, split real and imaginary).

StefanPochmann on April 23, 2015, 10:59 p.m.
<blockquote> You just do black magic </blockquote> ftfy But your input parsing is even better. Wow. I hated it and you made it look so easy.

veky on April 24, 2015, 4:09 a.m.
<blockquote> <blockquote> You just do black magic </blockquote> </blockquote> You can't fool me, I know you understood what I wrote there. ;-P <blockquote> I hated it and you made it look so easy. </blockquote> I hated it too, that's why I made it look so easy. ;-]] (That sentence is probably deeper than it looks at first sight.;)

StefanPochmann on April 25, 2015, 1:15 p.m.
<blockquote> I know you understood what I wrote there </blockquote> I did, but your program doesn't look like it's doing that. I'll try actually doing what you wrote there, maybe then I'll understand the program. <blockquote> That sentence is probably deeper than it looks at first sight </blockquote> Too deep for me to see. (Hey nice, that sentence has words alternating between vocals 'o' and 'e'... maybe I should turn that into a mission... nah)

veky on April 25, 2015, 10:45 p.m.
<blockquote> Too deep for me to see. </blockquote> Hm. I wanted to say that I struggled with it precisely because that format wasn't really looking as something that would be directly usable, and yet it seemed so close. Of course I evaled and unpacked it at two levels, and then worked with complex numbers created from that, but it frustrated me and I was sure there was a better way. I remember thinking "if only Python had a flag like HP49G, to interpret 2-tuples as complex numbers", and then it hit me. :-D And on another level, I like to hide that mental process, at least at first, because inducing the "how the hell did he do that" effect is a powerful teaching technique. :-D That's the "made it look easy" part. When Sherlock works on a case for a long time, the solution always looks so incredibly simple, precisely because he invested a lot of effort into its sublimation. I see you're different, and you like <a href="http://www.checkio.org/mission/fizz-buzz/publications/StefanPochmann/python-27/shortest-52-kill-me-now/share/2c00da177f66ce76430037ef425a0665/#comment-29444">disclosing your struggles</a> even unprovoked. :-) But that's ok, one more Veky here would probably be too much. :-)

StefanPochmann on April 25, 2015, 11:26 p.m.
Ok, makes sense. Most of the time I don't explain, either, there I mostly did because those expressions are so twisted that I feared people might think I found them with a program and take that away as the lesson.

hrvoje on June 19, 2015, 10 p.m.
Daaaaamn, that's clever. And short. :)

veky on June 20, 2015, 1:42 a.m.
Not as short as it could be. As I said, be sure to read the comments. ;-)

R2R on Nov. 27, 2021, 7:47 p.m.
Creative AF! :) Perfect jib!

veky on Nov. 27, 2021, 9:26 p.m.
Not really perfect (new tests invalidated it), but good enough. :-)

veky on Nov. 28, 2021, 4:48 a.m.
It seems the "invalidation" was temporary (the same solution now passes all tests). So maybe it is perfect. :-]

Ice Base

0 %

You ClassRoom Demo is Ready With the new tool CheckiO ClassRoom, you will be able to analyze your students' progress, have your own leaderboard, change missions order on the map, and so on.