Module talk:ET

Ratios for ed-phi, ed-e, and ed-pi

local common_suffix = {
	['3/2'] = 'f',
	['1618033988749895/1000000000000000'] = 'ϕ',
	['2'] = 'o',
	['2/1'] = 'o',
	['3'] = 't',
	['3/1'] = 't',
	['2718281828459045/1000000000000000'] = 'n',
	['3141592653589793/1000000000000000'] = 'π'
	
}
local common_ratio = {
	['f'] = rat.new(3, 2),
	['ϕ'] = rat.new(1618033988749895, 1000000000000000),
	['o'] = 2,
	['t'] = 3,
	['n'] = rat.new(2718281828459045, 1000000000000000),
	['π'] = rat.new(3141592653589793, 1000000000000000)
}

I had to undo a recent change (see above) since it was resulting in several templates breaking; specifically, the rat module was taking a very long time to calculate these large ratios and resulting in the interpreter terminating with signal 24. If these changes are necessary, might I suggest hardcoding them instead? Ganaram inukshuk (talk) 01:15, 22 January 2024 (UTC)

I just did that change to ensure that the irrational numbers are fully accurate, so it's not necessary. I've put it down to 6 decimal digits now.

CompactStar (talk) 01:52, 22 January 2024 (UTC)

Those aren't significantly more accurate than before. For pi its accuracy is actually worse. Definitely not worth the load time. Hardcoding is the way. FloraC (talk) 06:29, 22 January 2024 (UTC)

Update

Using various approximations for pi etc. does not make any sense so they are commented out. We should rewrite the math so the equave is not required to be a rational number, this restriction doesn't make sense anyway. I will take a look at how to clean up this mess soon.

EDIT: I uncommented again since pages like 9edϕ were broken.

Sintel (talk) 20:31, 1 April 2024 (UTC)