The APL Orchard

« first day (306 days earlier) ← previous day next day → last day (2349 days later) »

2:37 PM

Hey @Adám! Could you give this answer a read to see if I can golf it more? I've tried everything I know, but I couldn't get shorter than that.

Adám

2:55 PM

@J.Sallé OK, there's a lot you can do:

J. Sallé

3:31 PM

@Adám I found one more thing that I did in the REPL but not on TIO which is replace the first ×10* for E

Adám

To get you started:
https://tio.run/##ddPNattAEMDx@z6Fbk5CIvZbuw@wlz3l5kPpwdDiGhvL0IQQSk8BHwIOhRD6Cr31UIKhx/pN9CKOkghpRjMFY8P87OEvrTzbrC4@3c5W9fzYPDxdXjbbH/ZYb9qPb83ub/s6b3a/rJRGp3@/lTvsQwxVUqYwPrnC68onJQvnky2sCjbFQn2YzKfXN8Wk2f1p7u9Us3183/Tx8LN9P@yb3fOZ/n48atvulfJkspxPinpzakqTtBHDePqFna@vl5@/snJ1s1gvF//B5WJVX9VrHjOcCi@dhmW2DD4mbQF0bVSGOmqoj1kKCylnPBfeVApWujJW9u0XPXSVVIZKaqiSWQorKWc8F8ZEdMq@tCq@3uUBukoqQyU1VMkshZWUM54LH6NCN1OVIbZUQek6GRpCGUSl3F6YynjGIFzQRuOD98El7aH0J08IHj3B0dnTvfjwiWcMQjvj0fGHlto/mYPStTI0tDKIWrm9sJXxjIG2qlJq@3oZtJUh@AwQHD0DdC9@BohnDEKp4HGqUSppDaAvHQsMHduokyzFmWPOeN4@w97I0S2…

J. Sallé

@Adám wut I don't even know where to start there :p

no idea what ⍺⍺ means

Adám

@J.Sallé In order to avoid having to take apart ⍺ due to APL functions being limited to only two arguments, I made it an operator, which can have up to four arguments/operands. ⍺⍺ is the left (and in this case, only) operand.

@J.Sallé ⍺⍺ is the + in 2+/list where the / corresponds to op.

J. Sallé

Oooooooh I see

So, you index the "operand", which is the measurement unit in this case, with 1['gWuw '⍳⊃1↓⍺⍺] to decide which of the divisors to use?

Well maybe "index" is not the best word here

But I think I got it?

Adám

@J.Sallé Yes, I look up the second letter, because the first is ambiguous, in the string, and then use that to select factor.

J. Sallé

3:44 PM

@Adám I see. And 40032E¯15 is part of the gravitational binding equation which you then use as the dividend for the measurement unit, then takes that ×⍺÷⍵*2?

Adám

@J.Sallé I just moved the terms around, swapping × ←→ ÷ to be able to put ⍵*2 on the far right, but …40032E¯15×⍵×⍵÷⍺×8987E13… works as well.

J. Sallé

@Adám I see, I see. I'll post that then :)

Adám

@J.Sallé Go ahead. However, you may be able to golf a byte or two by refactoring 40032E¯15 and/or its exponent with the other terms.

J. Sallé

@Adám Okay, I'll try to do that

Adám

4:41 PM

@J.Sallé The 1↓⍺⍺ explanation line belongs above ⊃.

J. Sallé

@Adám oh, okay. I'll edit it

7 hours later…

Uriel

11:28 PM

@Adám regarding the vector normalization, my attempts actually came pretty close - the only difference is that I attempted the *.5 on the right, rather than commuting

but yea, I really need to practice trains seriously for sure

at least for the 2018 dyalog competition ;)

Adám

@Uriel Yes, that's a rule of trains, that f g A → A g⍨ f

@Uriel That isn't about golfing. You can win that with tradfns only too.

Uriel

btw, how much of the English language did I disband with my second message?

Adám

@Uriel Maybe (when we're done with the functions' marathon) we should have a lesson on trains?

Uriel

@Adám uh, I pretty much know most of the stuff about it. I just need to be less lazy when feeling urge to solve with a dfns :P

Adám

@Uriel Not much, you just need a couple of periods: but yea, I really need to practice trains. Seriously. For sure.

@Uriel You, yes, but there could be others who would benefit from a dedicated trains/tacit functions' lesson.

Uriel

11:34 PM

@Adám ugh, the original meaning was "for sure, I really need to practice trains seriously" with "seriously" adverbed to "practice trains"

@Adám oh sure, make them such session

Adám

@Uriel oh ok.

Uriel

@Adám lol, English is just so ambiguous

Adám

@Uriel That's why we have APL; to enable unambiguous communication of computational ideas.

Uriel

@Adám btw I thought of using your work on the Math lib as a basis for work on the eigenvalues challenge. are you ok with it (I figured you might want to take a go at it yourself)?

Adám

@Uriel You mean github.com/Dyalog/Math?

Uriel

11:39 PM

I started with an attempt to symbolify with complex numbers and solve with det(A-jI) but figured it won't handle complex matrices

@Adám yup, if it would come of any help

Adám

@Uriel Go right ahead. I did that as part of my normal work for Dyalog. It is Dyalog's repo, i.e. free-for-all.

@Uriel Can you use dfns.dyalog.com/n_det.htm for anything?

Uriel

@Adám this approach is toasted, but maybe other methods will need a determinant calculation. thanks!

« first day (306 days earlier) ← previous day next day → last day (2349 days later) »