The Nineteenth Byte

CMQ in python if I have a list of probs called probs I can sample an integer with random.choices(range(len(probs)), weights = probs)[0] This isn't beautiful or golfed but if I want to also return the probability of the number that was randomly chosen I can't think of an evenly vaguely golfed way to do it. Any suggestions?

e.g probs=[0.5, 0.5]; range(2).index(random.choices(range(2), weights=probs)[0])

image we have a list L instead of range(2)

slightly more compact :)

CMQ I can use random.choices(L, weights = probs)[0] to sample one element from the list L. But what if I also want to return the probability of the sampled element?

pxeger

item, weight = random.choices(zip(items, weights), weights=weights)[0]

graffe

oooh

I like it!

maybe [item, weight] = random.choices(zip(items, weights), weights=weights) ?

PyGamer0

@Ginger when did you change your name

graffe

@pxeger TypeError: object of type 'zip' has no len()

@pxeger seems that doesn't work sadly

pxeger

list(zip()) then

or use Python 2

graffe

3:58 PM

no to python 2 :)

item, weight = random.choices([*zip(items, weights)], weights=weights)[0]

fewer bytes :)

pxeger

you don't need the weights= btw, it can just be random.choices([*zip(items,weights)], weights)

graffe

I didn't know

thanks so much

I can't see a way to get rid of the [0] sadly

which is just ugly

pxeger

(item, weight), = random.choices(zip(items, weights), weights=weights)

graffe

ValueError: not enough values to unpack (expected 2, got 1)

you need the list(zip bit too

the problem is that choices returns a list containing a tuple

[(item, weight)] = random.choices([*zip(items, weights)], weights)

that works for example

pxeger

ato.pxeger.com/…

graffe

4:05 PM

ah the magic , ??

pxeger

it's equivalent to [(item, weight)] but one byte shorter

DLosc

4:21 PM

7

mathcat

wait where's the graph of Malbolge?

... oh lol

Neil

@pxeger doesn't , default to a tuple rather than a list, so it's ((item, weight))

pxeger

@Neil on the LHS of an = sign lists and tuples are completely equivalent

and ((item, weight)) doesn't work; you need a comma, because otherwise it's just got extra parentheses for grouping. ((item, weight),)

DLosc

Might the code be more readable if it were written as [[item, weight]] = ..., or is it always best practice to use tuples on the lhs?

Neil

@pxeger ah, so (item, weight), is actually a tuple of 2 all along?

pxeger

4:30 PM

@DLosc It's best practice to always keep the LHS simple enough that you don't need any brackets ;)

@Neil no, it's a tuple of 1

but you need the extra comma because otherwise Python thinks it's just an expression not a 1-tuple

DLosc

@pxeger Well, fair. So by that logic, indexing [0] on the rhs would be better?

pxeger

Yes, probably

for actual code, not for golfing obviously

DLosc

ofc

@Neil It's a 1-tuple which contains a 2-tuple.

Neil

so why did [(item,weight)] work but not (item,weight),?

pxeger

both work

Neil

4:32 PM

which version was complaining about the ValueError then?

pxeger

that was just (item, weight)

which is a 2-tuple, instead of a 1-tuple containing a 2-tuple

Neil

I'm confused because I don't see that version in the chat

DLosc

It was this one:

35 mins ago, by pxeger

(item, weight), = random.choices(zip(items, weights), weights=weights)

which errored because it was zip(items, weights) instead of list(zip(items, weights)), I believe

Neil

ah thanks it looked as if it was erroring on the assignment instead

DLosc

I am a little confused because that code produces a TypeError rather than a ValueError, so maybe graffe was referring to something else

Ah, probably was referring to an attempt to remove the [0] without changing the lhs yet: item, weight = random.choices([*zip(items, weights)], weights=weights)

pxeger

4:43 PM

@Neil I'm pretty sure that error was because graffe typed it in wrong

41 mins ago, by pxeger

(item, weight), = random.choices(zip(items, weights), weights=weights)

40 mins ago, by graffe

ValueError: not enough values to unpack (expected 2, got 1)

Neil

fair enough

pxeger

ato.pxeger.com/…

Sandbox Posts

5:01 PM

0

A: Sandbox for Proposed Challenges

Convert from Two's Complement to Decimal Take an input, and convert it from Two's Complement notation (binary where the first bit is negated, but the rest are taken as normal) into decimal. Input can be as a string, a list of digits, a number, or pretty much any other format which is recognizably...

Ginger

5:16 PM

regex be like [\"\'](\\.|[^"\'\\])*[\"\']

mathcat

5:36 PM

Do you speak regex?

I don't

DLosc

I do (though some regex flavors are still mystifying to me)

It's a very useful tool

Ginger

but hell to learn

regex is like javascript but more arcane and less mocked

mathcat

Regex is like the geometrical mean of PHP and JS

4

pxeger

I don't understand why people say regex is hard to learn

It's really easy to start from the basics and can be learnt perfectly in small chunks

DLosc

Regex is like a better version of search wildcards crossed with a simple golflang.

Ginger

5:47 PM

~~but is it turing complete~~

DLosc

~~depends on the flavor~~

~~chocolate~~

~~nice~~

~~stop~~

~~why~~

5:52 PM

CMQ: Opinions on "ungolfed" regex syntax like Python's verbose mode?

pxeger

pretty cool

mathcat

noice

DLosc

I never use it, but maybe I should

Some of the regexes in the Pip interpreter are getting... a bit long, shall we say

[A-Z]+|[a-ik-z_]|\$[][()$`'0-9]|\$[a-z][a-z0-9_]*

Bgil Midol

6:14 PM

@DLosc yes

mathcat

Retina Golfers: And I took that personally

ophact

@DLosc Interesting idea

emanresu A

6:48 PM

You can arrange a pickup of your sword in Rivendell between the hours of noon and 7:00 PM.

Ginger

the item is in transit to the destination

SegFaultPlus4

@emanresuA then the UPS building turns out to contain Alduin

oops

star the message about Alduin just to confuse anyone looking at the starboard

Ginger

according to my estimations it should arrive by 6 pm

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Remove unmatched brackets Inspired by @AviFS. Given a string containing of brackets, e.g. [xyz]543]abc[12345[def]hij[67890 Some of them, marked bold here, will be unmatched. ]abc[ [ ]543 12345[ ]hij xyz def [67890 Your challenge is to remove those unmatched bra...

Complex Irrational O-Fish with Ginger, Lyxal, and OLIMAR

ŤŇB Ğämę Čłůb

Transcript for