Timeline for draw ASCII function plots
Current License: CC BY-SA 3.0
26 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 6, 2015 at 23:19 | answer | added | blutorange | timeline score: 3 | |
| May 6, 2015 at 19:45 | comment | added | FUZxxl | @blutorange The scoring scheme is flawed anyway. | |
| May 6, 2015 at 18:48 | comment | added | blutorange |
I think numerics should count like strings, or somebody could take some data consisting of bits bbb... and encode that as a number 0.999bbb..., which counts as pretty much 0 (Log_256(0.999...)). Or perhaps, count numbers without the decimal point or other separators: 1.75 would count as 175, the rational number literal 9992345/10000000r in ruby would count as 999234510000000 (and not as 0.9992345) etc.
|
|
| May 6, 2015 at 15:23 | answer | added | Ewan | timeline score: 1 | |
| Feb 4, 2015 at 20:19 | history | edited | FUZxxl | CC BY-SA 3.0 |
added 26 characters in body
|
| Feb 4, 2015 at 20:18 | comment | added | FUZxxl | @MartinBüttner Good catch... | |
| Feb 4, 2015 at 18:56 | comment | added | FUZxxl |
@Ypnypn As described in the question, a string counts as one point per character it contains. For instance, the C string "a\n" counts as two characters, the C string u"aä", too.
|
|
| Feb 4, 2015 at 17:38 | comment | added | FUZxxl | The first behaviour (the "crappy" one) is correct. I'm on a trsin right now, let me amend the specificarion in a minute. | |
| Feb 4, 2015 at 17:35 | review | Close votes | |||
| Feb 5, 2015 at 16:45 | |||||
| Feb 4, 2015 at 17:30 | comment | added | Peter Taylor |
Or, indeed, am I guessing wrongly that the shaded pixels in the column should always extend to the x-axis? For all I know, the example image shows sin(x)sin(1000x) and the line does actually pass through all of those shaded pixels.
|
|
| Feb 4, 2015 at 17:28 | comment | added | Peter Taylor | The spec needs to explicitly state the criteria for shading a cell, because there are a number of plausible options and we can't tell which is intended. In particular, is the cheap and crappy "shade a pixel iff its midpoint is between 0 and the value at that x-coordinate" intended, or should the program attempt to compute the extrema of the image of the column and either shade the pixel iff its (centre-bottom / midpoint / centre-top) is between 0 and one of the extrema, or shade the pixel iff at least 50% of its area is? | |
| Feb 4, 2015 at 17:00 | comment | added | FUZxxl | @MartinBüttner No. It's correct like it is. Where un(der)specified, your solution is free to choose reasonable behaviour. This is an explicit design decision in crafting this challenge. | |
| Feb 4, 2015 at 16:59 | comment | added | Martin Ender | I like this question, but I think it could have benefited from a pass in the sandbox. | |
| Feb 4, 2015 at 16:59 | comment | added | FUZxxl | @MartinBüttner The horizontal line denotes the place where y = 0 on the cartesian coordinate system you are plotting on. The output format is intentionally a little bit fuzzy so you have more freedom for finding a solution. | |
| Feb 4, 2015 at 16:58 | comment | added | Martin Ender | @FUZxxl Of course, but that doesn't tell me where I should put it if y = 0 falls between two lines of the ASCII output. An example output would be helpful. | |
| Feb 4, 2015 at 16:57 | comment | added | FUZxxl | @MartinBüttner The horizontal line does not have to be exactly in the center. | |
| Feb 4, 2015 at 16:57 | comment | added | Optimizer | Also, it cannot be both code-golf and atomic-code-golf | |
| Feb 4, 2015 at 16:57 | comment | added | Martin Ender |
More importantly, how should we the horizontal line be placed if it doesn't fit exactly on a particular row like in your example? (e.g. if the function is a sine, but h is even)
|
|
| Feb 4, 2015 at 16:57 | comment | added | FUZxxl | @Optimizer Fixed that loophole. | |
| Feb 4, 2015 at 16:56 | history | edited | FUZxxl | CC BY-SA 3.0 |
remove tag code golf.
|
| Feb 4, 2015 at 16:56 | comment | added | Martin Ender | If your main concern is readability of the code, ask people to include an ungolfed version. If your main concern is golfing languages, then I guess you'll have to go along with scoring by tokens, but in my experience it doesn't really help a lot, because in most languages a function call is 3 tokens (function name plus parentheses) + 1 token for the first argument + 2 tokens for every further argument (because of a delimiter like a comma), whereas in CJam, say, it's just number of arguments + 1. | |
| Feb 4, 2015 at 16:52 | comment | added | Optimizer | It does not and should be clarified by the OP if he does not intend to let people abuse them, like you did with string and number | |
| Feb 4, 2015 at 16:52 | comment | added | FUZxxl | @MartinBüttner Hm… That counts as a standard loophole in my opinion. | |
| Feb 4, 2015 at 16:51 | comment | added | Martin Ender | Your definitions of how to count tokens are probably going to be abused as they are. E.g. by base64-encoding the code, naming a variable like this (which is a single token) and then getting a string representation of the variable name and base64-decoding it again. | |
| Feb 4, 2015 at 16:51 | comment | added | FUZxxl | @FryAmTheEggman I didn't know that tag existed. Thank you for the information. | |
| Feb 4, 2015 at 16:46 | history | asked | FUZxxl | CC BY-SA 3.0 |