python-discussion

[swift sparrow]

sure, but I'm only around for the next 15 min or so

[honest cradle]

meh that should be enough to finish up the classes

and im good enough to finish the rest i suppose

[lime sage]

im installing pycharm, what checkboxes should i put on?

[swift sparrow]

the ones that you want

[rare gazelle]

honestly, regarding the bool conversation:
what you think about:

if bool(x):
  ...

to be valid, but:

if x:
  ...

to be invalid

[lime sage]

but what are reccomended

[lime sage]

idk what they do

[swift sparrow]

I'd imagine the default ones are recommended

which ones are you unsure of?

[honest cradle]

https://discord.com/channels/267624335836053506/1502906292497547275

[spice hill]

timeit also accepts a string that it execs

[rare gazelle]

are those the same?

[lime sage]

the "crete ascociations"

[honest cradle]

i cant even use jetbrains stuff due to my trash hardware

[spice hill]

These are the same, as long as you didn't redefine bool

[latent lion]

def sort_the_inner_content(words):
    word=words.split()
    newlist=[]
    for i in word:
        j=i.split()
        x=sorted(j,reverse=True)
        newlist.append(x)
    return " ".join(word)
sort_the_inner_content("sort the inner content in descending order")

[swift sparrow]

what does the full thing say?

[rare gazelle]

yea but i dont like using a string... oh i see what you mean, to pass globals only when using a string

[spice hill]

exactly

[lime sage]

.py, .ipynb, .yaml, .yml, .md, .txt, .json

[rare gazelle]

i see

[swift sparrow]

Ok, so that updates your OS so if you double click a .py file (or any other of those extensions), it will open it in pycharm

[lime sage]

ok ty i will check .py

[bronze dragon]

j = i.split() just gives you back a list containing a single element (i)
since i is already a result of .splitting on words, further .splitting on it can only result in a one-element list
and because of that sorting j does not really do anything
finally, what you're returning is " ".join(word), and the word list was never modified, so you get back the input again

[latent lion]

var = bool(input("True or False?"))

[bronze dragon]

that may not do what you think

[latent lion]

, /
__/ CHAIR!!!
| |

[native coyote]

krillions must lear python

learn

[brisk gazelle]

That will evaluate to True or False depending on the length of what has been entered, assuming the builtin names haven't been futzed with.

Well, the righthand will evaluate to that.

The line as a whole is an assignment statement and doesn't evaluate to anything because it's not applicable.

[rare gazelle]

how is issuperset implemented? i assume it doesn't care if the lengths are the same

[quartz fulcrum]

hi everyone,
good morning!
i am back to learning

[rare gazelle]

say the length is greater or equal, is it just:

all(map(a.__contains__, b))

?

[latent lion]

If you input True Then the variable var will hold a boolean value of True

[bronze dragon]

https://github.com/python/cpython/blob/main/Objects/setobject.c#L2405

[edgy kraken]

Objects/setobject.c line 2405

set_issuperset_impl(PySetObject *so, PyObject *other)```

[rare gazelle]

C 🤮

alright let me C....

[brisk gazelle]

The same is true of entering in False.

[latent lion]

Yep

[bronze dragon]

bool("False") == True

[brisk gazelle]

Though variables don't hold the objects to which they refer.

[latent lion]

They do

[quartz fulcrum]

how?

[brisk gazelle]

They're named signposts. Not containers of those objects. They're proxies.

[bronze dragon]

nonempty strings are always truthy

[latent lion]

print("Welcome to python")

[rare gazelle]

i see... so its basically what i said

if the other is set though, it does issubset (i wonder why?)

[bronze dragon]

because it's enough to implement only one of {sub,super}set, and define the other using that implementation
when both are sets

[rare gazelle]

i see

would be nice if not NotImplemented would return NotImplemented

[brisk gazelle]

Wouldn't that be Implemented?

[rare gazelle]

its a special case that makes sense though

[brisk gazelle]

I mean, I know the Python correct evaluation, I'm just making a joke of it, but...

[edgy kraken]

Hey @lime sage!

Please edit your message to use a code block

```py
print('Hello, world!')
```

This will result in the following:

print('Hello, world!')```

[lime sage]

'''

[bronze dragon]

the correct Python evaluation being -- nothing, it's an error to use NotImplemented in a boolean context (since 3.15)

[brisk gazelle]

` not '

[lime sage]

how do i do a code blcok

[bronze dragon]

backticks, not single quotes. they're usually the key above Tab.

[brisk gazelle]

Look below esc, above tab, left of 1

On ~

[lime sage]

ty

[craggy trench]

the ```
~
`

key

[lime sage]

    def __init__(self, number, suit):
        if number > 10:
            number = 10
        self.number = number
        self.suit = suit

mycard = card(10, "daimonds")
print(mycard.number)
print(mycard.suit)``` im new to python is this good

[rare gazelle]

it's making it more explicit... i guess its fine... didn't know they changed the behavior

[brisk gazelle]

```py
...
...
...
```

[bronze dragon]

yep, a good use of classes.

[brisk gazelle]

snake_case your mycard variable and PascalCase your card class name.

!pep8

[edgy kraken]

PEP 8

PEP 8 is the official style guide for Python. It includes comprehensive guidelines for code formatting, variable naming, and making your code easy to read. Professional Python developers are usually required to follow the guidelines, and will often use code-linters like flake8 to verify that the code they're writing complies with the style guide.

More information:

• PEP 8 document
• Our PEP 8 song! :notes:

[brisk gazelle]

Mind your spelling.

diamonds vs daimonds

[terse mauve]

I would do:

self.number = min(number, 10)

[lime sage]

ok

[spice hill]

it's ai-enhanced diamonds

[half pewter]

ow ow ow caffeine migraine 🙁

[quartz fulcrum]

i want to use codex and claude's specials...

can anyone help me use these

[half pewter]

@honest cradle not a bot, just a regular here who appreciates seeing someone who's making the effort.

[half pewter]

what are "specials"?

[quartz fulcrum]

i heard claude is better for coding help

[dry pike]

oh same! I almost posted about it in off-topic

drinking a 473ml redbull at 8pm was not a good idea

[half pewter]

It's decent, though it can fluctuate. They're under no obligation to maintain any specific model for the user. I have found that a github subscription offers pretty reasonable access to a selection of models as well as github

[half pewter]

My synapsis were just feeling a bit frayed so I decided to have a day off. Hopefully it's not to persistant

[dry pike]

honestly I have no idea what you mean

[vapid mural]

Where can i learn filesystem UI/UX design? I want to add some style and better experience using it. Maybe there is nothing to think about and somewhat copy windows design

[half pewter]

we were talking about caffeine migraine? For me it's normally a withdrawal symptom

[pallid garden]

filesystem UI/UX?

[bronze dragon]

wdym? like if you were building your own OS, how would you organize the filesystem hierarchy?

[dry pike]

no I mean the words you said

[pallid garden]

like a file explorer?

[vapid mural]

I do got a video sample that i could show but if im allowed to upload it somewhere here or i can post on yt and share a link if im allowed

[bronze dragon]

you can share links yeah

[half pewter]

spun out I guess? third eye blind wrote a whole song about it

[vapid mural]

gime me a sec, im not sure about how its called its sort of file explorer and sort of file system

[quartz fulcrum]

i don't want to stay behind in tech things...
Can someone explain how codex works...or give a n example

[haughty crane]

is dijkstra really most efficient

[bronze dragon]

no, see A*

well, also depends on your goal

[haughty crane]

to find the shortest distance between 2 nodes

[haughty crane]

ty

[bronze dragon]

A* is better if you just care about the path between two specific nodes
Dijkstra will give you the shortest path between one specific node and all other nodes

[vapid mural]

https://www.youtube.com/watch?v=6uJLMzxQJNM

Here it is

It does work similar to file explorer but it has more functionality that fits for my app desires

[bronze dragon]

it's a bit confusing where one is at any point

like am I looking at boxes or categories (and what's the difference) or questions

file explorers have a location bar to tell you where you are at all times

[pallid garden]

usually

[brisk gazelle]

Instructions unclear, accidentally used supermassive black hole search.

[pallid garden]

the good ones do, at least

[vapid mural]

Well as it, its not FileExplorer app its Leitner System app, im allowing to create directories to sort things out and i create directory with suffix .box to know if you can run it

[bronze dragon]

funny thing, I first started the sentence with "most", then deleted it because I felt like it was too much hedging.

[vapid mural]

Its just a prototype at this stage but i want to explore my options forwards

[pallid garden]

anyway @bronze dragon is right, you should include a location bar

[vapid mural]

Noted

[rare gazelle]

one interesting thing
if not iterable: is useful also when handling iterators
if iterable: means that maybe the iterable is empty, maybe not
but if not iterable: means that the iterable is empty for sure

if not iterable:
  ...
else:
  ...

could be a reasonable pattern

[vapid mural]

I also for quick prorotyping use menu bar to do the work but i feel like it can get annoying to move mouse there so maybe adding some shortcuts or like right click dropbox windows have might be good

[granite wyvern]

I feel I'm missing some context here.

[half pewter]

i wasn't aware that iterable was an attribute of anything by design

[rare gazelle]

we talked about bool coercion before

[granite wyvern]

oh right

I like bool coercion

is this a desire featured of iterables you're descriing?

[rare gazelle]

iterator is always truthy, filled iterables are always truthy as well, so you can't know by if iterable if it's full or empty

but if not iterable is only right for empty iterables, so you can know by that the iterable is empty for sure

[granite wyvern]

Right. The truthy thing is for containers like lists and sets.

[rare gazelle]

and thus this pattern could be useful also for iterators

[granite wyvern]

The tricky bit is that on mahy iterables you can't know if it's empty without calling next(). Because the iterable may be a generator whose yielding is ependent on stuff.

So it's not something which can be implemeneted in general.

[rare gazelle]

the thing is it doesn't matter

[vapid mural]

So is there a good way to learn UI/UX without just searching what works and copying? Copying ofc with the understanding whys and whats

[rare gazelle]

you can still gain the knowledge of when it's empty for sure

[half pewter]

I'm not really part of this conversation but num iterables is when someone usually brings up len()

[rare gazelle]

thats the thing, you dont need to check len

[granite wyvern]

Not without trying to consume something from it.

[rare gazelle]

of -> when it's empty for sure

if not iterable would only enter if it's empty for sure

the else would mean, maybe its empty, maybe not

[tame vapor]

huh ? wdym ?

how would u "copy" UI/UX

[rare gazelle]

its the equivallent to the pattern:

if isinstance(iteralbe, Collection) and not iterable:
  ...
else:
  ...

equal to

if not iterable:
  ...
else:
  ...

[granite wyvern]

That'd be nice. But it's a magic test which can cannot work without causing the iterable's cogs to turn.

[tame vapor]

other than just looking at what is there and making it on ur own app

[vapid mural]

What kind of question is that? By copying...

[rapid ridge]

Anyone using napari ?

[tame vapor]

UI / UX is a general term , you said " i wotn copy without understanding " , what is there to understand is my question

like , "this website has a navbar , they make it collapse when the screen resizes" , like thats just a fact , u just make the same thing in ur app or u dont 🤷

[gaunt jolt]

Hii cuties

[rare gazelle]

you basically dont need the isinstance check... in other words

[vapid mural]

Well why the specific element is in this place and not this? Why did they place it there? Why does this element important?

shit i repeated xD

[peak relic]

Why do you ask?

[bronze dragon]

the elementaryOS HIG is a good and practical read about some fundamentals of UX design (even if one doesn't always agree with their conclusions): https://docs.elementary.io/hig
you also have books and blogs. I have read, for example, Bruce Tognazzini's blogs on design a lot. this post is good to start with: https://asktog.com/atc/principles-of-interaction-design/. Bruce was one of the first UI designers at Apple way back in the day. I like this a bit more than the eOS HIG.

[half pewter]

Welcome to the python server, you'll probably get a more agreeable response if you choose intellectual attributes to compliment though.

[vapid mural]

Thank you

[rapid ridge]

I want to know how to start contributing in it, what do i have to learn or understand

[gaunt jolt]

Im Rust Developer

[peak relic]

Reading the contributing guide would be a good first step. https://napari.org/stable/developers/index.html

[half pewter]

What brings you to this neck of the woods?

[gaunt jolt]

Ai ml

[half pewter]

we have 2 channels where they're discussed extensively https://discord.com/channels/267624335836053506/366673247892275221https://discord.com/channels/267624335836053506/1492906383748632746

[peak relic]

I'm pretty sure you can do ML in Rust. https://www.smartcorelib.org/

What interested you to do it in Python?

[half pewter]

this channel is appropriate for any general topics though

I can't imagine rust has a library that's as mature as pytorch, it really is a very nice library

[peak relic]

I guess if they wanted to, they could mix Python into their Rust with PyO3.

And that should give them the added advantage of using pytorch... I think

[scenic finch]

Wow never heard of this before

[simple willow]

Can I use my Education PyCharm for Personal Non-Commercial Projects or is it limited to just educational work?

[half pewter]

pycharm has a free teir, I doubt anyones going to care what you do with python that's non-commercial

[scenic finch]

Probably? I can’t imagine even if they had a rule how they would enforce it

[half pewter]

i'm sure jetbrains has some nice features in their pro model but a hobbyist is not part of their target paying demographic, definately respect any comercial contracts though

[simple willow]

I was just wondering as I have a student lisence and it has all the pro apps

which is just sat there

[slow rivet]

JetBrains IDE licenses for students are strictly for non-commercial educational purposes

[minor plover]

Heyayyaya wsup my gayss andd straightttttt kidsss

[simple willow]

so if I were to upload the code I made personally like a personal project to github I plan to make no money from it, I would be breaking the terms

[half pewter]

not appropriate

[young flicker]

Why not vscode

[minor plover]

Mah bad 😔😔

[simple willow]

I don't like VSCode

also a code editor is a personal preference

[young flicker]

Cursor

Neovim

Straight up Python's IDLE

[half pewter]

If I had pycharm pro access I'd probably want to at least try it but tbh I could see it being an issue down the line when I had to switch or pay

[simple willow]

a code editor is a personal preference

[slow rivet]

Maybe? That is from the "can I use it for work" faq, but the language does suggest it.

But also you can argue any personal non commerical project is educational as you're practicing python

Either way I highly doubt they will care

[young flicker]

Im giving you alternatives to PyCharm if you're worried about pricing or licensing

[simple willow]

would it be worth reaching out to someone at JetBrains to further query?

[half pewter]

doesn't cursor consume tokens?

[simple willow]

I am not worried about either, I am happy with the free tier. I am just wondering as I have a student lisence for it laying about

[young flicker]

Im sure you can write code without needing to use AI

[half pewter]

thought that was cursors whole gimmick though

[young flicker]

They do

[simple willow]

I have used the free tier for months

[tame vapor]

then thats just vscode

[young flicker]

https://lp.jetbrains.com/pycharm-for-students/

[nimble isle]

does random.choices with weights work with decimal weights? E.g. random.choices([1,2],[0.5,1]) makes 2 twice as likely as 1?

[tame vapor]

!d random.choices

[edgy kraken]

random.choices

random.choices(population, weights=None, *, cum_weights=None, k=1)```
Return a *k* sized list of elements chosen from the *population* with replacement. If the *population* is empty, raises [`IndexError`](https://docs.python.org/3/library/exceptions.html#IndexError).

If a *weights* sequence is specified, selections are made according to the relative weights. Alternatively, if a *cum\_weights* sequence is given, the selections are made according to the cumulative weights (perhaps computed using [`itertools.accumulate()`](https://docs.python.org/3/library/itertools.html#itertools.accumulate)). For example, the relative weights `[10, 5, 30, 5]` are equivalent to the cumulative weights `[10, 15, 45, 50]`. Internally, the relative weights are converted to cumulative weights before making selections, so supplying the cumulative weights saves work.

[young flicker]

No need to contact jetbrains lol

[slow rivet]

If you care about it sure, you can always ask their support.

[tame vapor]

The weights or cum_weights can use any numeric type that interoperates with the float values returned by random()
yes

[nimble isle]

(that includes integers, floats, and fractions but excludes decimals)
not decimals?

[simple willow]

doesn't answer my question...

[nimble isle]

also

Weights are assumed to be non-negative and finite
Then why are we allowed negative weights?

[scenic finch]

You should use the license if you have no reason to save it for later

[simple willow]

Personal, non-commercial projects that are not for profit and are purely for learning or experimentation are allowed. Yeah I am guessing I can

this just came from the AI bot

[scenic finch]

👍

[tame vapor]

but u are not using decimal , u are using float

[nimble isle]

oh yeah right

[simple willow]

and I can upload the code to github as long as I use the built in tool

[peak relic]

you can upload to github with any tool that has git. not limited to pycharm

[scenic finch]

I really don’t think they would care, and I think the ai is also hallucinating on this one

[tame vapor]

i dont think jetbrain will care at all unless ur project starts to make some money or gets popular or something

[simple willow]

It probably is

[jagged belfry]

No

[sacred charm]

what can this mean ? Step 8
Now, you want to create a string that displays the employee's age.

Start by creating a variable employee_info and assign it the result of concatenating:

the full_name variable.
(a string consisting of the characters is preceded and followed by a space.)

first_name = 'John'
last_name = 'Doe'
full_name = first_name + ' ' + last_name
address = '123 Main Street'
address += ', Apartment 4B'
employee_age = 28

[jagged belfry]

free educational licenses are to be used strictly for non-commercial educational purposes.

So, so long as you aren't making money with it or doing it for work (and you are a student), you may use the student version

Once you become not-a-student you can use their community version for similar non-commerical purposes, but it doesn't come with the Pro features.

[peak relic]

like how you concatenated the first name and last name for the full name, do the same for the employee info but to concat the age too

[jagged belfry]

The license is exclusively there so that companies have to pay and open source does not

[tame vapor]

also they say u should have a space before and after ur string

[sacred charm]

what is the is doing here

[peak relic]

full name is age

"John Doe is 28" is what I assume the expected output

[young flicker]

What are you working with here? Overall goal of the code?

[sacred charm]

ah ok

[peak relic]

Is this from freecodecamp?

[sacred charm]

freecodecamp employee info generatot

[peak relic]

Yea I figured

[sacred charm]

yeah just want a certificate

[young flicker]

Employee info generator?

[sacred charm]

yeah

[young flicker]

You wanted to generate fake data?

[sacred charm]

um not really ?

[peak relic]

Its part of the lesson on how to handle string data.. Basic stuff

[young flicker]

Ohh got it

[simple willow]

If I planned to make money, I would just use something like Zed or VSCode

[jagged belfry]

Then I would assume you're fine

[sacred charm]

employee_info = full_name + 'is' + employee_age

this should not work right

as the employee_age is a int

[peak relic]

needs spaces around is

[jagged belfry]

The full license is here https://www.jetbrains.com/legal/docs/toolbox/license_educational/

[sacred charm]

yeah i think they didn't mentioned to add the employee_age yet

[jagged belfry]

IANAL but to me only this line is important:

[peak relic]

which step are you on atm?

[jagged belfry]

(B) You may not:
(vi) use the Product for any commercial purposes.

The rest of it is boilerplate ( you may not share with others, you may not disassemble, you may not re-sell, etc )

[sacred charm]

its the step 8

[young flicker]

Have you learned formatted strings? It'll prevent type errors like that

Or you can just str(employee_age)

[sacred charm]

employee_info = full_name + ' is ' + str(employee_age)
print(employee_info)

this dont seem to work

[young flicker]

Yep like that

[peak relic]

oh ok. Dont add employee_age yet. Just full name and then the is with spaces around it

[sacred charm]

ah ok

[young flicker]

Type error? Or?

[sacred charm]

system dont accept it

[young flicker]

The platform?

[peak relic]

Just follow the instructions. It says to assign full name and the " is " and thats it

[peak relic]

freecodecamp

[young flicker]

Damn

[sacred charm]

employee_info = full_name + ' is '
print(employee_info)
``` this worked

thankfuly

[peak relic]

👍

[sacred charm]

tysm

[peak relic]

Just make sure to follow the instructions and you'll be fine

[sacred charm]

ah ok i was thinking beyond ✅

they should have introduced the f strings for that ..

[peak relic]

slowly... but surely...

[sacred charm]

Concatenating many strings using + and converting numbers using str() can get messy and hard to read.

Python 3.6 introduced f-strings to solve this. By adding the letter f before the opening quote, you can put variables and expressions inside replacement fields represented by curly braces {}. For example:

they just did it now lol

at step 13

[young flicker]

Lol

Yes, thats why I asked haha

Also at scale, for things like making full name, or full address components, pandas is pretty good and will scale beyond formatted strings once you get to that point

[sacred charm]

i hope so , i am just doing ts to get the certificate

[swift turtle]

But what will you do with this certificate.

[young flicker]

Ya for the certificate alone idk

[sacred charm]

idk but my mom wants that

[bronze dragon]

come on

[young flicker]

Certificates have arbitrary value to others, but if its for yourself and getting a feeling of accomplishment, then its good

[swift turtle]

Yeah I guess.

[sacred charm]

yeah it will look like aprove to smthin idk

[peak relic]

Maybe to HR ppl, they care

[young flicker]

Very few certificates they care about tbh

[bronze dragon]

certificates for basic courses are an anti-signal if anything

[swift turtle]

Lol true.

[sacred charm]

freecodecamp certificate is not good ?

[young flicker]

For a job? No.

[swift turtle]

"Introduction to Python" esque-certificates would probably render an anti-signal.

[sacred charm]

i cannot imagine going for an interview with that also lol

[peak relic]

its fine to have it

but it won't get you past the interview

[sacred charm]

yeah its not a sin ..

[swift turtle]

It doesn't hurt your application if you have something stronger to off-put it.

Like everyone and their mother probably has that same certificate, you need something to differentiate yourself from the next person.

[sacred charm]

OooO

what can that possibly be ?

[swift turtle]

I'm still figuring this out myself twin. 😔

[young flicker]

It wont get you to the interview in the first place tbh in this job market. Entry level positions are dwindling, hiring is stagnant, too much uncertainty cause of AI

[bronze dragon]

good projects, genuine contributions, prior actual experience (internships and such also count)

[swift turtle]

Agreed.

[ocean ridge]

have you ever done internships?

[young flicker]

Find a job role you want, match what the job role generally asks.

[sacred charm]

I have made some contributions in github and made some projects ..

[bronze dragon]

yeah

[young flicker]

Also if youre trying to go the data route - kaggle datasets will do nothing.

[swift turtle]

Certificates are like a coughing baby compared to the hydrogen bomb that is real-world experience.

[ocean ridge]

oh so you guys get placements often?

[young flicker]

Everyone and their moms and their moms mom have done kaggle projects

[swift turtle]

Lol.

[bronze dragon]

I didn't get them through the placement system, I got it outside. I have friends who did that though

[simple willow]

Certs can be good to get a start for learning, but it is not 'Get a Cert, get a job!'

[ocean ridge]

Oh alright. You are free to do freelancing there? or are there any regulations? I know some friends of mine who do freelancing while in college

[young flicker]

Get a degree, get several certs, do several projects =/= get a job, but its a step forward

[bronze dragon]

no regulations against it.

[ocean ridge]

alright. you are in iiit hyderabad right?

[bronze dragon]

do other colleges generally prohibit it? how do they enforce it?

[bronze dragon]

no. the University of Hyderabad.

[ocean ridge]

Oh alright. How did you get into it? JEE?

[bronze dragon]

yep

[sacred charm]

Define employee_code as 'DEV-2026-JD-001'. After that, create a variable department and assign it the slice of employee_code from index 0 to 3. Then print department to the terminal.

[ocean ridge]

whats the requirement

[sacred charm]

employee_code = 'DEV-2026-JD-001'
department = employee_card[0:3]
print(department)

ii am doing it right yeah ?

but why it dont accept

[young flicker]

0:2 possibly?

If it means 3 elements

[sacred charm]

no luck

[swift turtle]

Eh.

[bronze dragon]

look at your variable names, employee_code, or employee_card?

[peak relic]

which step?

[ocean ridge]

employee_code and employee card

[sacred charm]

16

[ocean ridge]

you accidentally wrote card

[sacred charm]

ah jesus

[ocean ridge]

employee_code = 'DEV-2026-JD-001'
department = employee_code[0:3]
print(department)

this should work ^

[ocean ridge]

ha we all make typos 😭

[sacred charm]

yeah it did

[bronze dragon]

you need to pass 12th I guess? if you mean what the percentile cutoff is generally, I wouldn't know... you can see statistics about that online. the national admissions system (JoSAA) publishes what was the last rank that got into any particular university.

[sacred charm]

yeah good luck finding the terminators 🙁

[ocean ridge]

Oh kk

[scenic finch]

Are there multiple IITs

[bronze dragon]

yeah.

23 I believe.

[scenic finch]

Are they all as good

[bronze dragon]

no, of course. the older ones have more established infrastructure, deeper alumni networks, more funding, better professors, etc.

[scenic finch]

Ah I see

When people talk about making IIT is the goal a specific one?

[young flicker]

@sacred charm wait how old are you? You said your mom wanted you to do it?

[sacred charm]

ah he is into online freelancing things ...

i am 15

[bronze dragon]

varies. they generally mean "any one of the old IITs", but they may be fine with the newer ones. or they may be aiming for a specific one due to personal factors like how close it is to them, whether they like the city, maybe particular areas that a single one specializes in, etc.

[young flicker]

Ah gotcha

Showing you numpy and pandas would be overkill for you I guess

[sacred charm]

its not like that i have learned multiple libraries like tkinter and pyqt5 i think i will learn them too one day

[young flicker]

Numpy/pandas are data libraries

You'd need to understand a lot of statistics, linear algebra to some extent, ideally have an idea of databases and SQL

[sacred charm]

ultra nerdy stuff ..

[young flicker]

You only need the python to learn GUI frameworks really

[young flicker]

Its where Python has its most competitive advantage compared to other programming languages

Python is strong for scripting and data

[sacred charm]

but the C and C++ talk directly to the hardware and are considerd fast ?

[young flicker]

The majority of the libraries in Python like NumPy and Pandas run under heavily optimized C/C++/Fortran code

Python just calls them in a way less boilerplate, safer, and user friendly way compared to those languages

[sacred charm]

thats why it is easy to learn , while being fast ..

[young flicker]

And thats why people dont use C/C++/Fortran instead of Python for things in the data ecosystem

[scenic finch]

Fortran?

[young flicker]

Except for very niche circumstances, custom approaches, high volume and high performance needs, or research grade things

[young flicker]

NumPy uses BLAS/LAPACK for the linear algebra functionalities which is from Fortran

[sacred charm]

i cannot get to understand this :

[scenic finch]

Interesting

[sacred charm]

pow(): raises a number to the power of another or performs modular exponentiation.

[scenic finch]

You’re right

[wide elm]

i have a unfinished github i found thats a good idea if anybody wanna help finish it using ai or smth

[sacred charm]

oh whats it ?

[grave tree]

Numpy also provides a nice way to call fortran subroutines from Python code which can be handy

[wide elm]

calculates crypto tax, there is not a single free site that does this so its kinda crazy imo

and its python code too

[bronze dragon]

I guess what you don't get is the "modular exponentiation" bit?

[sacred charm]

ah crypto stuff ...

[sacred charm]

yeah i dont understand the pow() function fully

[wide elm]

wym

[bronze dragon]

pow(base, exp, mod=None)
    Equivalent to base**exp with 2 arguments or base**exp % mod with 3 arguments

    Some types, such as ints, are able to use a more efficient algorithm when
    invoked using the three argument form.

see the base**exp % mod bit.

[sacred charm]

i dont know much about crypto

[bronze dragon]

that's what "modular exponentiation" means.

[wide elm]

how do i found somebody that does any channel here to find help?

[grave tree]

Like the tax you'd owe the gov't for various crypto transactions or something else?

[sacred charm]

create a channel in #1035199133436354600

[jagged nacelle]

Cryptocurrency or cryptographic?

[wide elm]

yes basically all countries have crypto tax (governments) and somehow there is not a single free site not kidding

most countries have crypto tax*

[sacred charm]

so if i give pow(2,2) it will be 2**2 ?

[wide elm]

currency

[bronze dragon]

yep

try it and see!

[grave tree]

Well taxes can get very complex, so that makes sense. Have you filed your own taxes before?

[sacred charm]

ah well let me try then

[wide elm]

crypto no cuz its expensive buying these sites and i found a github that needs a little polishing

i have deadline tho

i gotta file my taxes latest june smth

[sacred charm]

ah thanks its just a way to do exponential calculations or it have any other options ?

[bronze dragon]

the mod option is there.

[wide elm]

and thing is the group that made the github was like 5 years ago but they are crazy coders

[bronze dragon]

it lets you calculate base**exp % mod more efficiently than just doing the exponentiation then the mod

[sacred charm]

we could have been done it using 2**2 but why we did the using pow ?

[wide elm]

they work for crazy startups

[sacred charm]

whats the speciality

[grave tree]

At least from the US perspective, the tax code is a nightmare and it's not straightforward to calculate the tax you'd owe on crypto. The tax prep software can handle it, but it depends on what the rest of your tax form looks like for how much you'd actually owe

[hybrid nebula]

As said, mod

[sacred charm]

% mod meaning the remainder ?

[unreal vigil]

a local project with git can still use the files from gitignore, right? just that it doesn't look at the changes there

[hybrid nebula]

yes, a lot more efficiently

[bronze dragon]

yes

[grave tree]

Yup

[bronze dragon]

gitignore is just about what gets staged (and eventually committed/pushed)

[sacred charm]

like 2 to the power divided by remainder of mod ?

[bronze dragon]

the files themselves still exist

[hybrid nebula]

challenge: calculate 2 ** 1000000000000 % 3

[grave tree]

The project can use the files, git just doesn't track any changes to them

[wide elm]

in sweden its fairly easy i think just that my python isnt that good and no turtorial that teaches me how to make this (understandable)

but the github looks good tho

[sacred charm]

Oh 1 ?

[final hollow]

I might just switch to codeberg at this point because GH has lost a lot of my trust

[hybrid nebula]

pow(a, b, c) = a ** b % c but you never actually have to calculate a ** b

[hybrid nebula]

Yep

[sacred charm]

ah i get it first the exponent gets remainded* by the exponent right ?

[hybrid nebula]

Go ahead!
join us

[hybrid nebula]

what?

(a ** b) % c

[sacred charm]

first it do exponent % mod right ?

[hybrid nebula]

no

[sacred charm]

then why that ?

[hybrid nebula]

wdym?

[sacred charm]

never calculate? ?

a**b never calculate it ?

[hybrid nebula]

You get the result another way

[sacred charm]

whatever , i got the idea ...

[fiery yarrow]

why do i keep running into weird quirks of the typing system

[final hollow]

Gonna make a new set of operators

<(*-*<)
(^*-*^)
(>*-*)>

[sacred charm]

everyone have cats except me ...

[final hollow]

(v'-'v)

[fiery yarrow]

i wish fix was here 😔

[sacred charm]

freecodecamp is actaully fun and simple

[wide elm]

is it best python turtorial?

i had a little python but forgot a lot

[gleaming knoll]

even languages which allow defining operators usually dont allow having ( ) in them

[brittle merlin]

I finally understand all the AI generated async code in my codebase 😃😃😃

[sacred charm]

i dont know if if you have any prior experience in python even little its best

[young flicker]

The math or esp Numpy have a lot of extra math utilities if you wanted more math stuffs for your own use

[sacred charm]

that will be def overkill

[young flicker]

Like uh

Idk, pi maybe?

[sacred charm]

pi = 3.14

thats what i know xD

[gleaming knoll]

🥧 = yummers

[sacred charm]

what is this ? λ

[young flicker]

Lambda. Could mean a lot of things

[sacred charm]

a lot of thing ?

[young flicker]

Means something in math, something in stats

[sacred charm]

multinational symbol

[young flicker]

And its the symbol for lambda (the Greek letter for L)

And lambda functions are like one time/single use function

Anonymous functions

[sacred charm]

final_bill = running_total / num_of_friends
print('Bill per person:', final_bill)

each_pays = round(final_bill)
print(f'Each person pays: {each_pays}')

and the heading : Use the round() function to round final_bill to two decimal places and assign the result to a new variable named each_pays.

Finally, use print() to display the string Each person pays: followed by a space and your each_pays variable.

With that, the bill splitter workshop is complete.

it dont work ...

[gleaming knoll]

i think you're supposed to round to 2 decimal places, but by default round rounds to an integer

[sacred charm]

how to do that ?

[gleaming knoll]

!d round
it has a 2nd parameter for the amount of places to round to

[edgy kraken]

round

round(number, ndigits=None)```
Return *number* rounded to *ndigits* precision after the decimal point. If *ndigits* is omitted or is `None`, it returns the nearest integer to its input.

For the built-in types supporting `round()`, values are rounded to the closest multiple of 10 to the power minus *ndigits*; if two multiples are equally close, rounding is done toward the even choice (so, for example, both `round(0.5)` and `round(-0.5)` are `0`, and `round(1.5)` is `2`). Any integer value is valid for *ndigits* (positive, zero, or negative). The return value is an integer if *ndigits* is omitted or `None`. Otherwise, the return value has the same type as *number*.

For a general Python object `number`, `round` delegates to `number.__round__`.

[young flicker]

^

[fiery yarrow]

lambda you're kind of a typing enthusiast, right?

[sacred charm]

ah thanks i didn't paid attention on the time i was learning that

[gleaming knoll]

did you know that kinds are the types of types

[young flicker]

You could also round the formatted string so you only round what you print and not the actual values, but its probably not what youre being asked for

[gleaming knoll]

and i am not kind

[sacred charm]

yeah , they are asking for ..

[gleaming knoll]

death
a death sentence

[fiery yarrow]

why are the two returns treated differently by pyright

@dataclass
class Foo:
    x: int

class FooCollection(NamedTuple):
    foo1: Foo
    foo2: Foo
    foo3: Foo
    foo4: Foo

def order_foos(foos: list[Foo]) -> FooCollection:
    match foos:
        case [Foo(x=1) as f1, *f, Foo(x=4)]:
            # Expected 4 positional arguments
            return FooCollection(f1, *f, None)
            # no squiggles
            # return FooCollection(*[f1, *f, None])

[young flicker]

Mhm just use the round function

[sacred charm]

it can just round 5.89 to 6 ?

that will be not accurate

[gleaming knoll]

my assumption would be that since list has a single type for all elements, you get a union and then you unpack that union into all params
oh wait thats another thing no thats not what you're asking
list doesnt have a size in the typesystem so it just assumes it will match, sort of an optimistic thing to not cause errors in "real code" and cause people to rewrite to check length

[light ore]

https://x.com/i/status/2053388745554932067

[sacred charm]

dont click random links

Socrates ~

[young flicker]

You choose how you want to round it

[sacred charm]

oh yeah ..

[gleaming knoll]

there could be a special case in the typechecker for unpacking list displays
but
there just isnt so

for tuples the size is known in the type so that shouldnt happen

[young flicker]

Round(whatever, 2) - rounds to two decimal places

[fiery yarrow]

hmm switching to exploded tuple still has the same complaint as the no-list version

on-hover says that the match wildcard f is list[Foo]. i wonder if pyright thinks that FooCollection is getting only 3 values (or maybe even 2 for an empty list?), or some unknown giant number

[sacred charm]

what can this possibly mean ? : In Python, every value has an inherent boolean value, or a built-in sense of whether it should be treated as True or False in a logical context. Many values are considered truthy, that is, they evaluate to True in a logical context. Others are falsy, meaning they evaluate to False.

[gleaming knoll]

you can do bool(x) on almost all values in python
and things like if and while do it implicitly

[velvet trout]

When we can't?

[bronze dragon]

logical context meaning if x

[gleaming knoll]

when __bool__ = None or raises error

[bronze dragon]

NotImplemented, or Numpy arrays, for example

[sacred charm]

why this is true ? ```py
print(bool('Hello')) # True

[velvet trout]

That's sad

[fiery yarrow]

numpy arrays i think don't like being booled

[bronze dragon]

all nonempty strings are truthy

[velvet trout]

Non empty strings are true

[quartz fulcrum]

Bhardwaj told this in morni g

[velvet trout]

Empty strings are false

Even str() -> makes an empty string

[gleaming knoll]

strings are sequences of characters, a sequence is truthy if its not empty

[sacred charm]

ah ok so evry occupied string is truth ?

[velvet trout]

Yeah, its the truth of the universe

[sacred charm]

and this ```py
print(bool('')) # False

[velvet trout]

The ultimate reality

[sacred charm]

is false right

[velvet trout]

Yes

[bronze dragon]

"occupied" is a strange word to use there

[quartz fulcrum]

yes

[sacred charm]

my minds remember simple words ..

not augmented assignments ...

i will never remember that ...

[bronze dragon]

is "non-empty" not simple enough?

[sacred charm]

it confuses me just ,,

[quartz fulcrum]

what else is false

[bronze dragon]

[sacred charm]

in non or empty idk

[sacred charm]

0 or False

i think

[velvet trout]

empty -> False
is not empty -> True

[sacred charm]

print(bool(False)) # False
print(bool(0))  # False
print(bool('')) # False

[tame vapor]

my_str = "False"

[young flicker]

0 is false, 1 is true

[sacred charm]

childish concept

[tame vapor]

how is that childish 💀 , thats just how computers work

[sacred charm]

nah i mean i know it since i was born

😀

[gleaming knoll]

ok so uh
it seems like typecheckers assume that a *args of types of variable sizes fits all the missing args right away
so if you have any arguments after it - they're extra

[young flicker]

Binary is all 0s and 1s

[velvet trout]

[sacred charm]

is_citizen = True
age = 25

print(is_citizen and age) # 25

[velvet trout]

Nothing is childish

[sacred charm]

this confuses me ?

[gleaming knoll]

def take_4(_0: int, _1: int, _2: int, _3) -> int:
    return _0

def unpack_list(xs: list[int]) -> None:
    take_4(*xs) # assumes that it will match, works

def unpack_fixed_tuple(xs: tuple[int, int, int, int]) -> None:
    take_4(*xs) # for tuples it can check exactly

def unpack_variadic_tuple(xs: tuple[int, ...]) -> None:
    take_4(*xs) # like with a list

def unpack_list_inbetween(xs: list[int]) -> None:
    if len(xs) == 2:
        take_4(0, *xs, 3) # too many? what

def unpack_fixed_tuple_inbetween(xs: tuple[int, int]) -> None:
    take_4(0, *xs, 3)

def unpack_variadic_tuple_inbetween(xs: tuple[int, ...]) -> None:
    take_4(0, *xs, 3) # doesnt work
    if len(xs) == 2: # this actually narrows the tuple type so it works
        take_4(0, *xs, 3)

[velvet trout]

I don't eat children in dish

[sacred charm]

you should , they are in your basment

[gleaming knoll]

even with a len check
lol

[velvet trout]

[fiery yarrow]

so for my example of

case [Foo(x=1) as f1, *f, Foo(x=4)]:
    return FooCollection(f1, *f, None)
``` it sees FooCollection as getting *6* values instead of the required 4?

[gleaming knoll]

more like 5
because f(0, *xs) works, it assumes len(xs)==3

[sacred charm]

911 , whats you emergency ? there are childs in a guys basment

[velvet trout]

*args: int or explicitly is Unpack[int]

How about **kwargs or wait is Unpack for **kwargs?

[velvet trout]

Thunder breathing - 11th form

Nuke this guy

[sacred charm]

age = 19
``` is this truthy ?

[sacred charm]

there is only 1 form ...

[gleaming knoll]

keep a repl open and do bool(x) on things
and non-zero integers are truthy

[sacred charm]

ah ok

[gleaming knoll]

so * of variable sized things is supposed to be the last argument for some reason
weird

[fiery yarrow]

i wonder if that's noticed and intended behavior

[sacred charm]

49 out of 526 steps completed ...

that pisses me off

[fiery yarrow]

regardless, this discovering of yours has greatly simplified the MRE:

def foo(a, b, c, d): ...
foo(5, *[10, 20], 40)

[gleaming knoll]

i love simple meal ready to eats

[velvet trout]

watch the trilogy movie i think he has 11th form self made

[velvet trout]

Non zero integers are True

[sacred charm]

like zenistsu only have 1 form

[velvet trout]

-1 -> True, 0 -> False, 1 -> True

[sacred charm]

thanks i am now completely broken its enoughtlearning today

[velvet trout]

Isn't that simple

[sacred charm]

i should work on my project now ..

[sacred charm]

it is

[velvet trout]

Yeah do that, i learnt much more from doing projects and doing something i like

[sacred charm]

anyone have um sya 1920x1080 screen ?

[sacred charm]

i am working on a youtube automator .. which requires the comment section screen coords

mine have 1280x1024 screeb

so i want the 1920x1080 screen coords

[native coyote]

guys, I dont get bottom up dynamic programming 🙁

[sacred charm]

bcz its the res most ppl have

[tame vapor]

isnt this against TOS ?

[sacred charm]

it dont work as a bot

which spams

it is for creators who want to automate replying their comments section

as you can see here : https://github.com/sarimgamerop-cloud/Intelli-Comment

[gleaming knoll]

do you get top-down dynamic programming? you start with the "big problem", and express its solution as doing something to the solution of a smaller problem, and that chains until you get to the smallest problem
bottom-up changes the order in which you start subproblems to start at the smallest problem, and have the solution to the next problem depend only on the solutions to already solved, smaller problems. then you chain that "next solution" until you get to the one you want

[bronze dragon]

wait, so you just reply to comments with LLM slop?

I think your viewers will notice, lol

[sacred charm]

i dont care , i just thought it will be fun, without thinking twice .. just learning

[gleaming knoll]

you could have fib(n) start with fib(n-1) and fib(n-2), that'd be top-down (usually recursive, with memoization)
or you could start at fib(0) and then fib(1) up until fib(n), thats bottom-up (usually iterative, with an explicit table for the results)

[native coyote]

okay that actually helps because I somewhat understand the top-down DP (like with recursion). Do you think I should try to get better at top down stuff and finding recurrence relations and stuff before I move onto bottom up DP

[native coyote]

so like solving fib(1) and then using that to solve fib(2) until fib(n) is that what you mean?

[gleaming knoll]

yes
well
you can use fib(0) and fib(1) to solve fib(2). all of the things you already solved, not just the previous

[dry yacht]

what way are you referring? The recursive or the iterative one?

[native coyote]

ok

[native coyote]

the iterative version I think bc its the bottom up dp approach correct me if im wrong

[wise yarrow]

you'll need to find the recurrence relations in both bottom up and top down, bottom up just requires you to also find an iteration order

[gleaming knoll]

yeah, in both cases you're expressing 1 solution as depending on others

[rare gazelle]

set's lt, gt, is actually not very correct, typing wise

its an interesting concept with NonImplemented, maybe it's type correct

[dry yacht]

hm?

[rare gazelle]

on eq when NonImplemented is returned no error is raised when doing ==

on lt for example, an error is raised on NonImplemented when doing <

[gleaming knoll]

you dont need to typehint that you're returning NotImplemented

[dry yacht]

>>> { 1 } < { 1, 2 }
True

[rare gazelle]

technically, lt accepts any object, just like eq

yet it's typehinted as accepting only AbstractSet

[gleaming knoll]

object has no __lt__, so no

[rare gazelle]

i'm talking about set's lt

[rare gazelle]

^

[gleaming knoll]

no, im saying that there is no reason for __lt__ to accept an arbitrary object

__eq__ accepts an arbitrary object because its part of object
object.__lt__ doesnt exist, so some_cls.__lt__ is free to typehint it as accepting whatever

[dry yacht]

Well... that's just a nature of the language, i.e. you can typehint it, but typehints aren't enforced

[rare gazelle]

there is set.__eq__

    def __le__(self, value: AbstractSet[object], /) -> bool: ...
    def __lt__(self, value: AbstractSet[object], /) -> bool: ...
    def __ge__(self, value: AbstractSet[object], /) -> bool: ...
    def __gt__(self, value: AbstractSet[object], /) -> bool: ...
    def __eq__(self, value: object, /) -> bool: ...

[gleaming knoll]

thats what i said, yes

[dry yacht]

When did lambda say there wasn't?

[rare gazelle]

i dont understand what u're saying

what i'm saying is that lt is accepting every object as well

[gleaming knoll]

then what you're saying is wrong

[rare gazelle]

< raises an error

but lt isn't raising an error

it might return NotImplemented though

but eq might return NotImplemented as well

the only difference is that on eq NonImplemented there is no error raise, on lt there is

[gleaming knoll]

NotImplemented is not reflected in the typesystem at all

[rare gazelle]

(when doing == and <)

[rare gazelle]

when you return NotImplemented in lt of set, the error raised if u use < automatically

[distant garnet]

does anyone know kivy?

[brisk gazelle]

Not for some time.

[gleaming knoll]

im not sure why are you mixing up runtime with typing

what are you actually saying?

[dry yacht]

def f(x: int):
    print(type(x))

f("abc")

prints:

<class 'str'>
```despite me type-hinting it as `int`.

Whereas now I get an error:
```py
def g(x: int):
    if not isinstance(x, (int,)):
        raise NotImplementedError

    print(type(x))

g("abc")

[rare gazelle]

because there is no actual error raising inside lt

[dry yacht]

yes there is

[rare gazelle]

as a flat function, it raises no error, it could be written exactly like eq

yet eq accepts all objects, lt accepts only abstract set

[dry yacht]

what is a "flat function"?

[rare gazelle]

< vs lt

[gleaming knoll]

__eq__ accepting object is the result of it being part of the object protocol

there is not much reason for "1" == 1 to typecheck, its just Liskov Substitution
object.__lt__ does not exist, so each class is free to define the argument type to __lt__ as specifically as it wants. typeshed defines it to be the one that will actually work at runtime, to prevent more errors statically

[dry yacht]

< is just synatical sugar for __lt__

[rare gazelle]

its not

well, you might think of it like this

but < is different than __lt__

it handles NotImplemented, and reflaction

reflection*

[brisk gazelle]

@distant garnet Is there a question forthcoming? 🙂

[rare gazelle]

on NotImplemented it raises an error (in set, for exmple)

[rare gazelle]

yea but you could define eq then to accept only abstractset

[distant garnet]

Dude i wanted to start kivy development like im in high school , my OOPS is weak though any tutorials or vids you can recommend

[gleaming knoll]

you cant, liskov substitution.

[rare gazelle]

then why not making lt accept all objects as well

since it technically does

[gleaming knoll]

object.__lt__ does not exist

[rare gazelle]

i mean, in set's lt

[brisk gazelle]

!beginner I'm well versed in the basic concepts. I can take you through them now if you like.

[edgy kraken]

Go-to beginner resources

Here are the top free resources we recommend for people who are new to programming:

• Automate the Boring Stuff — an online book (also available to purchase as a physical book)
• Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
• Python Programming MOOC 2026 course — text-based lessons with exercises
• Corey Schafer's YouTube playlist

For a full, curated list of educational resources we recommend, please see our resources page!

[brisk gazelle]

OOPS, as opposed to Kivy, whuch I haven't touched for a long while.

[gleaming knoll]

because its not useful. why allow code that will error in practice?

[rare gazelle]

it doesn't error, thats the thing

it returns NotImplemented

[gleaming knoll]

lt does not error, but < will

[rare gazelle]

yes, but lt isn't wrapping < or reflecting on it

its the other way around

relying* not reflecting

if you think of lt as a function for itself, why it needs to care about <

[gleaming knoll]

you're proposing to allow code that will error in 99.9999999% cases just because some single person might implement Cls.__gt__(set)

[gleaming knoll]

dunders are fundamentally related to their operators, it does not make much sense to consider them separately. people dont use lt as a function directly.

[rare gazelle]

if the error would get raised inside lt, then fine

but since it doesn't, and it never errors on any object

[velvet trout]

lazy import x

is x present in dunder dir of that module?

[rare gazelle]

so you say the typehints of lt should be the typehint of <

[gleaming knoll]

yes. the name is bound right away, its just that the value will only reify when accessed

[rare gazelle]

maybe its a special case with NotImplemented

its kind of weird

[gleaming knoll]

NotImplemented is assignable to all types, just like Any

[rare gazelle]

yea but maybe when you return NotImplemented you think about what would happen then

[velvet trout]

Oh nice

[rare gazelle]

and base your type hints on that

[gleaming knoll]

too much headache tbh

[velvet trout]

I guess i will copy this behaviour too in my PEP 562 implementation for pre 3.15 versions

[gleaming knoll]

i'd prefer people to not write insane dunders

[velvet trout]

currently i just impl dunder getattr and don't fill in dunder dir

[rare gazelle]

maybe it'd be better if NotImplemented wouldn't exist to begin with

[dry yacht]

Keeping it easy and readable 🐒 :

class Composition:
    def __init__(self, f):
        self.f = f

    def __rshift__(self, g):
        def wrapper(*args):
            return g(self.f(*args))

        return Composition(wrapper)

    def __call__(self, *args):
        return self.f(*args)


@Composition
def compose(*args):
    return args[0] if len(args) == 1 else args

[rare gazelle]

@gleaming knoll whats your take on this

[gleaming knoll]

maybe yeah
lhs could explicitly call the respective dunder of the rhs, if it wants to allow it to implement the operation

[rare gazelle]

it can get complicated

i've made a pipe class

its pretty long

say you want to connect pipes together, or disconnect, etc

[velvet trout]

Other than .po and gettext and json/yaml/toml, what else do y'all use for text translations? And do you keep key by english words or by a prefix + N where N is an integer starting from 0

[rare gazelle]

when you think about all the combinations its a big head ache

what if you wnat to add a function to the start, what if to the end, and all that bbullshit

i've also made pipes immutable, and hashable

[thorn mist]

i was trynna make a parabola jump but i fumbled ```py import pygame
pygame.init()
screen = pygame.display.set_mode((1280,720))
clock = pygame.time.Clock()
running = True
dt=0

player_pos = pygame.Vector2(640,360)
while running:
for event in pygame.event.get():
if event.type == pygame.QUIT:
running = False
fade = pygame.Surface((1280, 720))
fade.set_alpha(100)
fade.fill("purple")
screen.blit(fade, (0, 0))
pygame.draw.circle(screen,"red",player_pos,40)

keys = pygame.key.get_pressed()

if keys[pygame.K_SPACE]:
    if player_pos.y < 720:
        for i in range(1,100):
            x = i * 0.1
            y = x * x
            player_pos.y -= 300 * dt * y * 10
if keys[pygame.K_s]:
    player_pos.y += 300 * dt
if keys[pygame.K_a]:
    player_pos.x -= 300 * dt
if keys[pygame.K_d]:
    player_pos.x += 300 * dt
pygame.display.flip()
print(player_pos.x,player_pos.y)
if player_pos.x < 0:
    player_pos.x = 0
if player_pos.y < 0:
    player_pos.y = 0
if player_pos.x > 1280:
    player_pos.x = 1280
if player_pos.y > 720:
    player_pos.y = 720

dt = clock.tick(60)/1000

pygame.quit()```

[edgy kraken]

Hey @thorn mist!

Please edit your message to use a code block

Make sure you put your code on a new line following py. There must not be any spaces after py.

Here is an example of how it should look:
```py
print('Hello, world!')
```

This will result in the following:

print('Hello, world!')```

[dry yacht]

Like, a wrapper around this?
https://man7.org/linux/man-pages/man2/pipe.2.html

[rare gazelle]

its my own thing, i dont wrap anything

i was inspired by Mamoun, from this server, he also had an implemention for it (but mine is completely different)

[bronze dragon]

couldn't that lead to recursive spirals?

[rare gazelle]

(he called it pipe)

[gleaming knoll]

if both dont implement it yeah
maybe it would need some kind of context tracking

[rare gazelle]

i haven't published my implementation, but there's a thread of it at its early stages, that i opened in here

https://discord.com/channels/267624335836053506/1438841486111281245

[dry yacht]

Can you explain what a "pipe" in your context is (or how it differs from the

unidirectional data channel that can be used for interprocess communication
)

[rare gazelle]

composition basically

Mamoun called it pipe

out of respect, i used that name as well

[dry yacht]

ah, I see. Yeah, it massively confused me that you called that "pipe" 🐒

[rare gazelle]

that thread is no where near the final implementation, but it represents the idea

[rare gazelle]

uh its what he called it

[native coyote]

ok, thanks

[rare gazelle]

it gets pretty messy

getting it to work well with typehints

and all the possible combinations

[dry yacht]

hmmm, I guess if I were to name it and I had to include the word "pipe" I'd call it "pipeline"

[rare gazelle]

well... thats fine i guess

he called it pipe, so i called it pipe

i guess its the answer to this python question that sometimes pop, of something recently i'm proud of, or maybe, was really fun to work on

[autumn forge]

it would be cool if you could partially apply each piped function

like P = pipe; P(a, 1, 2) | P(b, 2) | P(c, 3), where a, b, and c all take 2 positional arguments

[rare gazelle]

not sure what u're saying

[dry yacht]

like, a "binding" pipe?

[autumn forge]

ah my example makes no sense, edited it

[gleaming knoll]

is that

c(b(a(1, 2), 2), 3)
# 
_0 = a(1, 2)
_1 = b(_0, 2)
_2 = c(_1, 3)

?

[autumn forge]

yes

[dry yacht]

aaaah

[willow elbow]

helloworld("print")

[autumn forge]

maybe you could do this with a decorator instead...py @piped def a(x, y): ... @piped def b(w, z): ... a(1, 2) | b(2) | ...

that would be nasty

[gleaming knoll]

but to get the actual value you'd need to (a(1, 2) | b(2)).value then
which kind of sucks doesnt it

[autumn forge]

hm, true

yeah that does kinda defeat the point

[gleaming knoll]

jast write ze fokin temporaries
python is not a place for your awesome composition pipelines

[granite wyvern]

When I've done things like this each pipeline stage is an object wrapping a function. Typically with a mode indicating how it handles the stuff coming down the pipleine - one-to-many, batches etc.

[quartz fulcrum]

@granite wyvern can you help me in the thread..

[granite wyvern]

yes just reading it now

[rare gazelle]

you can use partial

in my version pipes can be connected and disconnected, combined in whatever way

but if you want to use a function with given argument you can pass a partial

[rare gazelle]

note that it'd work only with keywords

because partial can't do partial(x, ..., 1,2,3,4)

def partial_mod(function, *args, **kwargs):
    @wraps(function)
    def function_wrapper(*additional_args, **additional_kwargs):
        function_args = []
        function_kwargs = kwargs | additional_kwargs
        
        additional_args_iterator = iter(additional_args)

        for arg in args:
            if arg is not Ellipsis:
                function_args.append(arg)
            else:
                function_args.append(next(additional_args_iterator))

        function_args.extend(additional_args_iterator)

        return function(*function_args, **function_kwargs)

    return function_wrapper

you can use this for it

[bronze dragon]

now it can (since 3.14 I think)

[rare gazelle]

uh

[autumn forge]

yeah there's a new Placeholder object you can use there

[bronze dragon]

https://docs.python.org/3/library/functools.html#functools.partial

[rare gazelle]

every function i write eventually goes to the garbage i guess

[bronze dragon]

see the bit about Placeholder

[rare gazelle]

ew

well, you can use a different name i guess

Placeholder as _

from functools import partial, Placeholder as _
remove = partial(str.replace, _, _, '')
message = 'Hello, dear dear world!'
remove(message, ' dear')

[brisk gazelle]

That's a very dismal sentence.

[bronze dragon]

_ as a name is very liable to get overwritten
you're better off doing _ = Placeholder just before using it, if you really want

[rare gazelle]

you mean _ = Placeholder

its basically the same

[bronze dragon]

yeah yeah

[autumn forge]

or: your idea was useful enough to be added to the standard library (albeit in a slightly different way)

[rare gazelle]

x as _ or _ = x

[rare gazelle]

well... i guess what i did is still relevant for < 3.14

although i like the _ solution, i might use that as well

_ = ... is also neat not gonna lie

[velvet trout]

os.path.basename() faster than Path().name? how much?

[bronze dragon]

!d timeit

[edgy kraken]

timeit

Source code: Lib/timeit.py

This module provides a simple way to time small bits of Python code. It has both a Command-Line Interface as well as a callable one. It avoids a number of common traps for measuring execution times. See also Tim Peters’ introduction to the “Algorithms” chapter in the second edition of Python Cookbook, published by O’Reilly.

Similar names: ipython.timeit

[velvet trout]

ipython?

[rare gazelle]

i dont like combbining os paths with Path

[velvet trout]

i ain't combining, am picking either

[rare gazelle]

i'm saying that you should pick what you already use, if you use Path stuff then use Path

[velvet trout]

class Command(Parameter):
    def __init__(
        self,
        name: str | None = None,
        *,
        aliases: str | Sequence[str] | None = None,
        parse_mode: ParseMode | None = None,
        subcommand_required: bool = False,
    ) -> None:
        self.name = name or Path(sys.argv[0]).name

Path() will probably be used for root-level command which is usually once

[bronze dragon]

IPython has special magic for it (%timeit for a single line, or %%timeit for a whole cell)
but the timeit module exists independently of IPython

[velvet trout]

Ooohhh

[rare gazelle]

(Path is normally considered slower than os)

[velvet trout]

Ruff says use Path and i like pathlib, so ema use that unless i find performance issues

[rare gazelle]

yea i think thats the correct route to go with

[velvet trout]

Obviously i can shh ruff

[gleaming knoll]

(you wont find performance issues)

[velvet trout]

Owo

[rare gazelle]

i dont use pathlib normally, but its just me i guess

its normally fine

[velvet trout]

I mean,... Os isn't even a built-in, you gotta import, so yeah i wouldn't find issue importing pathlib

[rare gazelle]

yea

i think that if you dont face performance issues (and as lambda said, u won't, normally at least) picking pathlib is better for all cases

[gleaming knoll]

there is no need for this "normally"

[velvet trout]

I don't like having such a verbose name with keyword or containing in it, although its the most concise and clear enough name i have got, do you guys can suggest me better name over specifier_or_option ?

def remove_option(self, specifier_or_option: str | Option, /) -> Option:
        if isinstance(specifier_or_option, str):
            try:
                option = self._specifier_to_option[specifier_or_option]
            except KeyError:
                exc_message = f"Option with specifier {specifier_or_option!r} not found"
                raise ValueError(exc_message) from None
        else:
            option = specifier_or_option

        try:
            self._options.remove(option)
        except ValueError:
            exc_message = f"Option not found: {option}"
            raise ValueError(exc_message) from None

        for specifier in option.specifiers:
            del self._specifier_to_option[specifier]

        return option

[velvet trout]

makes sense

[rare gazelle]

spec_or_opt

[velvet trout]

still or in it

[gleaming knoll]

there is no application that you would run that does anything meaningful, such that when profiled, you'd find Pathlib().name taking such a significant amount of time such that even if os.path.basename was faster, it would be worth switching to it

[rare gazelle]

altohugh spec might refer to spec..

[velvet trout]

Yeah

I also have marked it as pos-only

[bronze dragon]

what does a specifier look like?

[velvet trout]

1. '-a'
2. '--foo'

[bronze dragon]

just say option then

[rare gazelle]

i think that if you go fully pathlib, or fully os, there could be a difference where hundreds of thousands of files are involved

[velvet trout]

Is this my chance to use NewType ?

[bronze dragon]

option: str | Option, that is

[rare gazelle]

you must have this or?

maybe make 2 functions?

remove_specifier remove_option

[bronze dragon]

fundamentally the "specifier" also just represents an option
so there's no user-visible need to distinguish it from an Option object

[velvet trout]

hmmm, but option: Option is also used in lots of places. Won't that knowledge collide with the user?

[rare gazelle]

why not just calling it option yea...

[bronze dragon]

in those other cases, why not also accept a specifier string there?

[rare gazelle]

make it the same in all those places

[velvet trout]

i am afraid they would think its about concrete Option than a str lol... Is it okay

[rare gazelle]

either all supporting str and not, or just Option in all

[bronze dragon]

you should aim for a consistent interface unless there is a strong reason to only take Option in some places and take str | Option in some

[velvet trout]

I don't think those places have knowledge of specifiers but i will check it out

[bronze dragon]

the typehint exists for that

[velvet trout]

What from typing would be appropriate for making a typename Specifier instead of str

type Specifier = str ?

NewType(str)?

[bronze dragon]

eh, if you make it a new type it would be more confusing IMO

[velvet trout]

And would be great if it can talk about prefixes such as - and --

[rare gazelle]

a class that wraps or subclasses str

[velvet trout]

str seems too generic lol

[bronze dragon]

like I hover over it, I see Specifier | Option, that doesn't immediately tell me that yeah Specifier is actually a str

[velvet trout]

you're right

[rare gazelle]

yea if its just str i'd use str

[bronze dragon]

unless I already know that issubclass(Specifier, str)

[velvet trout]

specifier is always a str everywhere

[rare gazelle]

then maybe 2 different functions

remove option and remove specifier

[velvet trout]

I feel like i am making a bad designed code lol

[velvet trout]

but its cool to be remove without double funcs lmao

[unborn frigate]

mates is tehre a chat here where I can ask for an opinion? I dont want specifically help with something but to understand a concept for jupyter notebooks

[rare gazelle]

the naming is less cool

[bronze dragon]

or, just take Option, and have a user-visible way to get_option_by_specifier

[cerulean ravine]

ask here

[rare gazelle]

you can also create a function for the common work if you want

[unborn frigate]

thanks mate, aight so Im new to python comin from C++, C# languages that I used for game dev.
I had this class that Im using Jupyter notebooks, but I treated it like a normal IDE. When i try to break my code in smaller pieces, nothing works not even if i try to boot like 4 lines of code. So I am confused

[rare gazelle]

def _remove(whatever):
  ...

def remove_a(x):
  ...
  return _remove(x)

def remove_b(y):
  ...
  return _remove(y)

[bronze dragon]

can you clarify what you mean by "nothing works"? do you get an error, or something?

[pallid garden]

each cell has to be a valid "file"

you cant split up one function into 2 cells, for example

[gleaming knoll]

well, they can rely on state from other cells since they are running in one interpreter, and they can even be ran in non-linear order

[bronze dragon]

each cell has to be independently syntactically valid

[unborn frigate]

yeah i get a huge error, so like

If i tried to run

#file will be provided please correct the right path to use
# used pandas documendation 
file_path = r"FAT - SF6 Frame Data.xlsx"

xls = pd.ExcelFile(file_path)
# check the sheet names and keep the ones with Normall in 
move_sheets = [name for name in xls.sheet_names if "Normal" in name] ``` just these lines I still get error if i press ctrl + enter

[cerulean ravine]

when I work in a notebook, I rely heavily on "restart kernel and run all code"

[bronze dragon]

please show us the actual error, just saying "error" doesn't give us nearly enough information to actually help you.

[pallid garden]

what is the error

[unborn frigate]

https://pastebin.com/XSG8wMyv its huge sorry dont wanna cover the whole chat

[pallid garden]

ImportError: Missing optional dependency 'openpyxl'. Use pip or conda to install openpyxl.

do you know what a dependency is?

[unborn frigate]

ye

[gleaming knoll]

^the error message itself is at the bottom of the stacktrace, always check it first

[unborn frigate]

I assumed that because it worked for my VS codium for another class

it woukld be all tehre

[hollow rampart]

Hm

[unborn frigate]

(im on cachyOS if it matters)

[pallid garden]

did your other class require you to open an excel file?

[unborn frigate]

no

[pallid garden]

every program is different

[unborn frigate]

ah right I did this thing before in windows

[pallid garden]

it doesnt matter

[unborn frigate]

now it makes sense why

[pallid garden]

you will get the same error on windows

[unborn frigate]

ye i mean i must have installed it there

[pallid garden]

why cant you install it on the machine you are currently using

[lime sage]

hi

[unborn frigate]

as shamefull as this is to save I didn't actualyl read the error because it was so huge I thought soemthign is broken

[bronze dragon]

always read the error

even if it's huge

[pallid garden]

i mean... it is scary

we are but veterans who are used to slaying scary monsters /jk

[unborn frigate]

just usually don't get errors so I was so confused even from 4 dialogue lines haha

[bronze dragon]

one of the most important skills is to be able to extract information from huge errors/texts even if you don't fully understand them.
learn to skim.

[unborn frigate]

but you aright

[pallid garden]

the most important attribute is courage to read the error message... /hj

[bronze dragon]

I get like a morbid curiosity. like I wonder what I screwed up to get that big of an error.

[pallid garden]

as my high school math teacher used to say for calculus problems: "take a deep breath and integrate"

[unborn frigate]

we got this ! thank you mates i split it on just my functions and I can see it working now that Im trying to load it cell by cell

[grim hatch]

Good morning

What's on the Agenda today

[unborn frigate]

this also helped a lot thank you

[rare gazelle]

😁

[velvet trout]

Where where else is if TYPE_CHECKING: used outside indenting typing-only import stmts?

do type checkers read or mutate this constant?

[gleaming knoll]

type checkers dont run your code. they just assume TYPE_CHECKING: Literal[True] even though at runtime its False, for the purposes of static analysis that means that the if branch is reached and else is unreachable

[velvet trout]

You got me

[rare gazelle]

[velvet trout]

type checkers never make use of this attr?

[gleaming knoll]

they do

[velvet trout]

how so and when

[rare gazelle]

^

TYPE_CHECKING is False at runtime, but True for typecheckers, a trick

[velvet trout]

i mean they just assume, but do they ever import it? Do anything with it other than blind faith that its True? lmao

[gleaming knoll]

no, static type checkers dont run python code at all.

[rare gazelle]

never thought about the else...

very weird

[velvet trout]

That is very vague at second part of the sentence, "but True for typecheckers, a trick"

[rare gazelle]

so you can do:

if not TYPE_CHECKING:

anywhere, and it'd be ignored for all type checkers?

🤯

[velvet trout]

it makes me wonder that ... do they set it to True? or the python itself?

[gleaming knoll]

yes

[rare gazelle]

🤯🤯🤯

[velvet trout]

Wonder if ruff yells at if not TYPE_CHECKING

or typecheckers

[velvet trout]

[rare gazelle]

maybeb it'd say this condition never enters

[gleaming knoll]

type checkers keep a mapping from qualified symbols to types
the symbol "typing.TYPE_CHECKING" is given the type of Literal[True] (its representation in terms of the internal structures used in the typechecker). that does not involve running python at all. its just internal state of the typechecker

[rare gazelle]

warn for that

Code is not analyzed because condition is statically evaluated as falsePylance

yea

if not TYPE_CHECKING:
    print("hello")

[minor plover]

HEYAYAYAYAYYAYAYAYYAYAYY MAH BOIIIS

AND GAYSSSS

WSSSSUP

[rare gazelle]

static type checkers dont run your code

they analyze it, statically

for a static type checker, TYPE_CHECKING is considered True at all times

but at runtime, TYPE_CHECKING is False

[rare gazelle]

yo bro, please no spamming

i know your ign is spammer and it might be hard for you

try to resist the temptation

[minor plover]

😔☝🏿 i am good human beinggg dw

[rare gazelle]

😁

[velvet trout]

Who said to run my code, i am just saying if it just imports from typing module to check the constant

[gleaming knoll]

"import typing" is "running python code". type checkers dont do that.

[velvet trout]

So a blind faith, it doesn't even bother importing TYPE_CHECKING, i see

[rare gazelle]

type checkers dont run your code

[velvet trout]

/Assumption

[rare gazelle]

TYPE_CHECKING is a constant, located in the typing module

[gleaming knoll]

you can imagine a typechecker as a function Graph[ModuleName, AST] -> Sequence[Error | Warning | Info]

[velvet trout]

i thought if a type checker is written in python so it could just import it and check it lol

[gleaming knoll]

there is no reason for it to check it, its defined by the spec that typecheckers should treat it as True

[velvet trout]

[velvet trout]

i see

[rare gazelle]

a static type checker isn't running any python code

it analyzes your file

[velvet trout]

What's the difference between "oh" & "i see" i use them interchangeable to say "oohhh is that so ..."

[gleaming knoll]

i mean, if it were to check it, it'd find that its false, which defeats the point

[velvet trout]

I thought they set it for themselves lol

[gleaming knoll]

so you did think they run the code?

[velvet trout]

Won't typing.TYPE_CHECKING makes the indented code inside it unreachable always? Or typecheckers/linters special case it?

[gleaming knoll]

the whole point is that its special cased

[velvet trout]

ign?

[gleaming knoll]

they do qualified_names_to_types["typing.TYPE_CHECKING"] = WhateverStructureTheyUseForRepresentingAType.Literal(True)
so when analyzing a program which has if TYPE_CHECKING: in its source code, the analyzer assumes that the branch is taken, even though at runtime its not

[velvet trout]

xD

Thanks guys

Now back to code

[autumn forge]

in game name

[rare gazelle]

in game name

[velvet trout]

How am i supposed to know that... How do you guys know that

[pallid garden]

too much water

[velvet trout]

i = too
g = much
n = water

[rare gazelle]

from games

[velvet trout]

Specifier = NewType("Specifier", str)

type OptionLike = Specifier | Option

then i can just do:
def def remove_option(self, option: OptionLike, /) -> Option: ...

What do you say?

I already have cliargparse/hints.py for OptionLike to be putten into

[bronze dragon]

why not give the user a way to get_option_by_specifier and then only use Options everywhere in your interface

[rare gazelle]

i think this is clean

or in Option's init accept a specifier

or a class method from_specifier

[rare gazelle]

there is strlike i think, reminds me of it

[bronze dragon]

PathLike also

[velvet trout]

I don't find it good to split tbh... it also makes easier for user since they're used to str and option concrete

Or maybe i am being biased lol... i gotta think more

[bronze dragon]

splitting what?

[rare gazelle]

yea PathLike

[velvet trout]

Two methods to remove option

[velvet trout]

How is that

[bronze dragon]

what I suggested would only have one method

[rare gazelle]

tell me, if there is options_or_specifiers, can it have 1 option and 1 specifier inside it?

[velvet trout]

This code i mean

[rare gazelle]

i see, 1 option or specifier, not an iterable

[velvet trout]

Huh

I don't get it

[rare gazelle]

i remembered it as remove_options my bad

[gleaming knoll]

Iterable[A|B] vs Iterable[A]|Iterable[B]
(those are not equivalent)

[bronze dragon]

remove_option(option_object)
# or
remove_option(get_option_by_specifier("--option"))

is what I mean

[velvet trout]

Does variance make it equivalent or something

[rare gazelle]

eitherway, what bharadwaj suggested is that it'd bbe only Option

[bronze dragon]

it's more orthogonal. it's plausible to me that users would like a way to go from specifier to option generally even outside the context of remove_option.

[gleaming knoll]

no, neither covariance nor contravariance nor invariance

[rare gazelle]

and there'd be a way to create an Option from a specifier

[velvet trout]

ohhh, get_option() only takes specifier so your design is good and possible but again remove_option can itself find option of that specifier since they're defined under same class