algos-and-data-structs

[fiery cosmos]

ill take a look @ the structure

[ionic pasture]

ok

[fiery cosmos]

thanks!

[lunar jacinth]

can someone help me understand how to utilize my vertex and graph function

for undirected graphs?

ive been stuck for quite a while and opened like 4 help channels with no responses

[restive grotto]

What are date structure and algorithms

[flat sorrel]

Well the idea is that the nodes would eventually reach an equilibrium where the nodes are more or less distanced according to the input subgraph. Btw this is just a random thought so I may be totally off xD

You construct the final (dense) graph according to the distance matrix

The number of dimensions you use should be consistent with the spatial constraints in the context that the graph represents.

It might also be worth considering nodes that have leftover degrees of freedom in their coordinates (due to not having enough connections)

[brave oak]

that is a very physics answer

that I do not have the background to appreciate

I have solved the problem in a much poorer way

but thank you for your input

I gotta say though

what you suggested sounds MEGA cool

I wouldn’t even know how to start doing that

🥴

[flat sorrel]

!e

import numpy as np
from scipy.spatial.distance import cdist

n_points = 10  # No. of nodes
n_dims = 2  # No. of spatial dimensions
n_iter = 500  # No. of iterations
damp_factor = 0.5  # Reduce velocity by this proportion per iteration
accel_factor = 0.01  # Ratio of force to mass (learning rate)

# Original distance matrix
dummy_coords = np.random.rand(n_points, n_dims) * 10
target = cdist(dummy_coords, dummy_coords)
print(f"Target:\n{target}")

# Simulate the input
prune_ratio = 0.2
idx = np.ravel_multi_index(np.tril_indices(n_points, k=-1), target.shape)
to_prune = np.random.choice(idx, size=int(prune_ratio * len(idx)), replace=False)

pruned_target = target.copy()
pruned_target[np.unravel_index(to_prune, target.shape)] = np.nan
pruned_target = (pruned_target + pruned_target.T) / 2
print(f"Pruned:\n{pruned_target}")

# Set up initial state
box_size = np.ptp(target) / np.sqrt(n_dims)
coords = np.random.rand(n_points, n_dims) * box_size
velocities = np.zeros_like(coords)

for n in range(1, n_iter + 1):
    print(f"[Iteration {n}]")
    
    # Shape: (n_points, n_points, n_dims)
    deltas = coords[np.newaxis] - coords[:, np.newaxis]
    
    # Shape: (n_points, n_points, 1)
    dists = np.linalg.norm(deltas, axis=-1, keepdims=True)
    diffs = np.nan_to_num(dists - pruned_target[..., np.newaxis])  # Ignore the pruned connections
    errors = np.abs(diffs)
    print(f"Train Error: {errors.sum() / 2}")
    
    # Shape: (n_points, n_points, n_dims)
    unit_vects = np.divide(deltas, dists, out=np.zeros_like(deltas), where=dists!=0)
    accels = -np.sign(diffs) * errors * unit_vects  # a = F / m where F = k * error
    
    # Shape: (n_points, n_dims)
    new_velocities = velocities * (1 - damp_factor) + accels.sum(axis=0) * accel_factor
    
    # Update state
    coords += (velocities + new_velocities) / 2
    velocities = new_velocities

[halcyon plank]

@flat sorrel :white_check_mark: Your eval job has completed with return code 0.

001 | Target:
002 | [[ 0.          8.55590774  8.34378067  7.30506343  9.69365725  7.79709852
003 |    6.2691179   4.87532748  4.17133523  2.8336265 ]
004 |  [ 8.55590774  0.          1.22205037  2.58555024  1.88326539  5.67348333
005 |    6.44558182  3.89974462  5.50053533  9.0027916 ]
006 |  [ 8.34378067  1.22205037  0.          1.53364889  3.06785746  6.63951194
007 |    7.18117649  3.48767616  4.84587458  8.41150492]
008 |  [ 7.30506343  2.58555024  1.53364889  0.          4.46539201  7.18684936
009 |    7.34988324  2.48494214  3.48275978  7.03677166]
010 |  [ 9.69365725  1.88326539  3.06785746  4.46539201  0.          4.9934119
011 |    6.24456377  5.42575124  7.14252204 10.52683554]
... (truncated - too many lines)

Full output: too long to upload

[flat sorrel]

@brave oak

The larger n_points gets, the smaller accel_factor needs to be to avoid divergence

A smaller prune_ratio also demands smaller accel_factor from what I have tried

For evaluation (I ran out of characters 😦 )

# Evaluate
deltas = coords[np.newaxis] - coords[:, np.newaxis]
dists = np.linalg.norm(deltas, axis=-1)
print(f"Result:\n{dists}")

diffs = dists - target
errors = np.abs(diffs)
print(f"Test Error: {errors.sum() / 2}")

[flat sorrel]

Ok I think I am done fixing bugs for the above xD

[fiery cosmos]

#discord-bots

[split nymph]

I found this line of code online for creating a hangman game I just wanted to understand it a little better if someone could explain pls

indices = [i for i, letter in enumurate(word) if letter == guess]

[fallen tulip]

do you guys have a turtle chat

[astral pelican]

Hi, I have two programs in my hand and I can't tell why one of them is working significantly faster than the other one. Can someone help?

Here's the slower one:

from collections import Counter
for _ in range(int(input())):
    n=int(input())
    w=list(map(int,input().split()))
    c,ans=Counter(w),0
    for i in range(2,2*n+1):
        cur=0
        for j in range(1,(i//2)+1):
            ad=min(c[i-j],c[j]) if i-j!=j else c[j]//2
            cur+=ad
        ans=max(ans,cur)
    print(ans)

And here's the faster one:

from collections import defaultdict, Counter

for i in range(int(input())):
    n = int(input())
    a = input()
    a = a.split()
    a = list(map(int, a))

    C = Counter(a)
    sum = defaultdict(int)

    for p in C:
        for q in C:
            if p <= q:
                sum[p + q] += min(C[p], C[q]) if p != q else C[p] // 2
    print(max(sum.values()))

Here's a sample testcase:

5
5
1 2 3 4 5
8
6 6 6 6 6 6 8 8
8
1 2 2 1 2 1 1 2
3
1 3 3
6
1 1 3 4 2 2

[kindred imp]

basically its looping through letter in the word, building a list of indices of the letters guessed that are in the word

indices = []
for i, letter in enumerate(word):
  if letter == guess:
    indices.append(i)```

[split nymph]

Yeah! I figured that out, thank you so much tho still

[fallen tulip]

ello

[swift ether]

Hello, I have one, probably, silly question here. I'm trying to divide two numpy arrays that are updated in real-time, and some of the values of them might be None. It's there a pythonic way to deal with this issue?

Thanks in advance

For example: array([5759.5, 3425, None, 10.8, 11.2, 39.13], dtype=object) / array([1231, 3212, 1243, 42.3, None, 39.13], dtype=object)

[agile sundial]

wait why are you using np to put multiple data types in

[swift ether]

Forget my question, you make me think jajaja I just need to turn the values into float

[sand kiln]

hello does anyone know How to write this function for this examples in the comment in one line code only?

[agile sundial]

doesn't seem that difficult

have you tried?

[sand kiln]

yes many times

[agile sundial]

what have you tried

[sand kiln]

my logic is if(((x*y)+z)*x:return true

something like that i think

[agile sundial]

isn't it supposed to return a string

[kindred imp]

Hoping someone can assist with this.
I am given a Class called Order, which is made up of ( id, symbol, quantity, price, side, timestamp), I want to store this order in an orderbook which is a list (not my choice) and when i store it in the orderbook, the list needs to be sorted by price and then timestamp.

There are two lists a buy list and a sell list
Is there an elegant way of inserting into the list the order so its ordered from insertion thats quicker than just append and sort

        my_list.append(order)
        my_list.sort(key=lambda k: (k.price, k.timestamp))```

can i use this
bisect.insort_right(my_list, (order.price, order.timestamp, order))

[brave oak]

you want a different data structure that inherently maintains an internal ordering

e.g. a self-balancing binary search tree

[fiery cosmos]

nightmares

[snow swift]

heapq?

[brave oak]

not fully ordered

you won't be able to get all the elements in sorted order in linear time

[half hornet]

if age>=18
print('test1')
elif age>=13
print('test2')
else
print('test3')

can anybody help me with this cuz the app I used for programming said that the syntax is invalid

[brave oak]

you need colons.

[half hornet]

ty

for the help

[rich cairn]

https://cdn.discordapp.com/attachments/778873265678450708/781769754746421268/unknown.png

guys can yall help me out with this error

[twin edge]

@rich cairn i guess you're missing closing parenthesis somewhere

[rich cairn]

hmm

what about the first error

the "else" one

[twin edge]

Fix that one first

[rich cairn]

finding

cant find it

[fiery cosmos]

@rich cairn
if
if >> elif (code):
if >> elif (code):
..
else:
code

[rich cairn]

@fiery cosmos ok

[onyx mist]

Hey guys. Probably simple solution but when I looped through this dictionary to minus the level from the capacity to create a 'fuel needed' it gave all these recurring values? simplest way to fix this inaccuracy?

Would rounding it to 2 decimal places make any further inaccuracies?

Looking through after trying that, it seems that rounding to 2 decimal places was fine

[slim cosmos]

hey guys is there a dedicated channel for OOP concepts?

[fiery raft]

does anyone know what is the max row count for parsing a csv with DictReader?

I have 1.4 million but when I do list(reader), len(list) I get approx 900k rows...

[fast flicker]

So I’m realizing some of my into-level CS students need more work on Python lists. Anyone have suggestions for some interesting problems that lend themselves to solutions using lists?

[agile sundial]

codingbat has a bunch of problems that explicitly use lists

[fiery cosmos]

@fast flicker like the methods of list or problems in which lists are involved?

[umbral matrix]

Hey guys, I'm Brazilian and I need help, with some codes, I specifically need help to put codes on the Cozmo robot, but I really didn't understand how to do this, nor how to use the robot's SDK properly ....

the code I intend to use on it is the voice control

[dark sinew]

hey guys , does anyone here uses openCV library on Pycharm mac os ??

[tacit lily]

hi i just heart that here can ask python module pandas thing and excel thing here . is it okay to ask here

[snow swift]

@dark sinew @tacit lily that seems more suited for #data-science-and-ml

@umbral matrix doesn't quite seem like the right channel; maybe get a help channel #❓｜how-to-get-help

[dark sinew]

@snow swift ok thank you , i asked there too but no respond till now T-T

[snow swift]

@dark sinew it'll also be better if you ask about your specific problem

since plenty of people know opencv

but noone can answer your question if you dont post it

[dark sinew]

so basically Pycharm can't get access to my mac's camera

for privacy reasons

i added some code on the pycharm Info.plist , it does work but when i close the app and reopen it it quits unexpectedly

so i have to edit the Info.plist every time

[ashen wharf]

def insertionSort(l):
    for i in range(0, len(l)):
        if l[i-1] > l[i]:
            for j in reversed(range(0, i)):
                if l[j-1] > l[j]:
                    l[j-1], l[j] = l[j], l[j-1]
                else:
                    break

    return l```

why does it not work

[fiery cosmos]

range(0, len(l)-1)

[ashen wharf]

still doesn't work

[fiery cosmos]

whats the error

[ashen wharf]

no error

but it doesn't sort

def insertionSort(l):
    sort = l
    for i in range(0, len(sort)-1):
        if sort[i-1] > sort[i]:
            for j in reversed(range(0, i)):
                if sort[j-1] > sort[j]:
                    sort[j-1], sort[j] = sort[j], sort[j-1]
                else:
                    break

    return sort

print(insertionSort([5, 4, 3, 2, 1]))

[fiery cosmos]

well, returns the same list

[ashen wharf]

nop

[fiery cosmos]

do you want it to show [1,2,3,4,5]?

[ashen wharf]

yes it's a sort

[fast flicker]

Problems that involve lists. I think they’ve got a handle on the basic methods, but I want to get them using them to solve real-world problems.

[fast flicker]

Well, the first thing I notice is that the syntax on your function call is wrong.

[ashen wharf]

ew

i forgot the parentheses when copying it

[fast flicker]

Okay.

There's another problem, though.

[ashen wharf]

yes?

[fast flicker]

Well, first maybe try adding a print statement to print sort after each loop. See if you can figure it out.

[4, 3, 5, 2, 1]
[3, 4, 5, 2, 1]```

That's what it does.

[ashen wharf]

ok

for some reason it doesn't check the last element

def insertionSort(l):
    sort = l
    for i in range(1, len(sort)):
        print(sort[i-1], sort[i])
        if sort[i-1] > sort[i]:
            for j in reversed(range(0, i)):
                if sort[j-1] > sort[j]:
                    sort[j-1], sort[j] = sort[j], sort[j-1]
                else:
                    break
        print(sort)

    return sort

[fiery cosmos]

class Tree(object):
    def __init__(self, entry, left=None, right=None):
        self.entry = entry
        self.left = left
        self.right = right


def insert(item, tree):
    if (item < tree.entry):
        if (tree.left != None):
            insert(item, tree.left)
        else:
            tree.left = Tree(item)
    else:
        if (tree.right != None):
            insert(item, tree.right)
        else:
            tree.right = Tree(item)
            
            
t = Tree(5, Tree(1, None, Tree(4)), Tree(7, Tree(6), Tree(8)))
insert(2, t)

print(t)
```So im trying understand binary search trees a bit better , when I use insert(2, t) is there a way to see the new updated tree ?

[ashen wharf]

@fast flicker

[fast flicker]

Just a sec.

Are you aware that range(4) returns [0,1,2,3]?

[ashen wharf]

for i in range(1, len(sort))

i gave it 5 numbers

it starts with 1 because i check i-1

[fast flicker]

You're only checking through i = 3.

[ashen wharf]

len(sort) = 5

because i inputted 5 numbers

[fast flicker]

range(0,4) => 0,1,2,3

[ashen wharf]

uh i said len(sort)

not len(sort)-1

[fast flicker]

I copied and pasted right from here.

[ashen wharf]

i changed it

#algos-and-data-structs message

[fast flicker]

So when sort = [2, 3, 4, 5, 1] and j = 0, is sort[j-1] > sort[j]?

[zenith pagoda]

In min-heap ff i will extract the 4, to which should I exchange it? to 90 ?

[shrewd cradle]

I think you bring up the last node, which is 87 in this case?

[zenith pagoda]

On the book CTCI it said that i swap it with last element in the heap which is the bottom most and right most. But the book illustration is quite confusing that's why I want to verify it.

[shrewd cradle]

Yeah I believe it's the bottom-right

[outer ledge]

is anyone available to help me with a python issue

[autumn mason]

ammm

guys

I need python experts

print(list(data[i] for i in range(len(data) - 1, -1, -1)))```
how does this function work? like I see the output, I just can't understand how it works in details

it gives as an output ['d', 'l', 'o', 'g'] but I would appreciate if someone could explain those -1,-1,-1 in the end

like the first one subtracts the length of the string, but the others do I can't comprehend

[brave oak]

range takes 3 arguments

start, stop, and step

in this case, stop and step are both negative

so the range outputs 3, 2, 1, 0

[quick tapir]

Hey! I can't figure out this math in my head so wondering if you guys could help me. Pretty much, I'm trying to calculate the rotation angle for a jump animation.
It gets three input variables; jumpLength, jumpAngle and jump. jumpLength says how many frames the jump animation should last, jumpAngle is what angle should be rotated from 0 (so for a 150* total angle, this would be 75* to say 75* down and 75* up from 0) and jump is what frame we're currently on.
By default, I want the angle to be 0, but must fall down to negative jumpAngle (e.g: -75) for a quater of the jumpLength. Then, when in the jump cycle, it will go up for half the frames (from -75 to 75) then back down for the last quater (75 to 0) where the process repeats.
https://repl.it/@CameronJohnston/Flappy-Bird Here's my current best attempt [34:], but I realize the math is completely wrong but I've already got a headache trying to figure it out
If anyone could give me a good pointer, that would be great!

[bold lake]

y dont u use https://en.wikipedia.org/wiki/Rotation_matrix

[quick tapir]

I'll have a look

Uh yeh I'm not smart enough to understand any of that sadly lol, but from what I an tell this is just for calculating the position of pixels after a rotation which isn't what I'm looking for since pygame handles that for me

[burnt hamlet]

Implemented a defaultlist :D https://github.com/Honno/coinflip/blob/da55d7ad8678cc76235bd4e190141d24d48d7c8c/src/coinflip/_randtests/common/collections.py#L123

Basically a glorifed wrapper of the defaultdict, as obviously it deals with a lot of the problems very well. I really wanted slices to be supported, and to first assume the slice range is for an infinitely-sized list as opposed to constrained by the "length" (highest defaultdict key) of the list

[quick tapir]

I figured my question out, and even managed to make it smoother 😮

https://www.desmos.com/calculator/z3kfscd4bc

[autumn mason]

Guys to assing certain type to argument is it suffciient to write
def reverse(self,x:int):?

if not how can I make x type int?

[agile sundial]

you can't

[fiery cosmos]

def reverse(self, x):
    x = int(x)
    ...

Not sure why you would want to do that though

The users of the function should be passing in the correct types

[agile sundial]

you can use static type checkers to emulate a statically typed language

something like mypy

you can't guarantee what type you get with vanilla python though

[autumn mason]

thanks guys

[autumn mason]

    def reverse(self, x):
        self.x = x
        return self.x[::-1]```

guys can somebody tell me why is it when paste integers into the method ```reverse```
I am getting ```TypeError: 'int' object is not subscriptable```

[agile sundial]

what does 4[0] give

[autumn mason]

4[0]?

I have gotten anything like that

[agile sundial]

i know

it's a hypothetical question

[fiery cosmos]

@autumn mason I think you want to take a number and reverse it

so there are two ways
Approach 1

def reverse(x):
  return int(str(x)[::-1])

Approach 2

def reverse(x):
  rev = 0
  while x:
    rem = x % 10
    rev = rev*10 + rem
    x //= 10
  return rev

[raven kite]

i have an array of integers, i need to split this to five sub array. And max sum between this arrays must be minimum. For example [10,10,20,30,30,10,20,15,10] to [10,10,10] [30] [30] [20 10] [ 20 15] max sum is 35 [20 15]

[fiery cosmos]

How big is the array

[raven kite]

its N

not improtant

it doesnt have limit

brute force is not accepted btw xd

[fiery cosmos]

Yeah thats what i was asking size for

What was your time complexity?

N^6?

[raven kite]

brute force currenlty

n!

im trying better

im trying to find better aproach

[fiery cosmos]

N factorial?

You can do n^6 by runnin 5 for loops and then other for loop to get the sums

i, j, k, l, m, n are the split indices

so 0..i

i..j

j..k

k..l

Aand l.. N

Oops 4 are enough

[raven kite]

but you are thinking like i need split array according the their sorting

no

i can choose one from start one from mid for one sub array

my aim is minimizing the max sum of sub array

[fiery cosmos]

Hold up

Lemme switch to pc

[raven kite]

i thought about sum of big array/5 and make subs array according to it

but it wont be always optimal

[fiery cosmos]

okay

so we have an array and we need to make 4 points to split the array into 5 subarrays

so we don't know which 4 points we are needed to split since we can't modify the array

[raven kite]

and

we dont need 4 splits

because we are not making sub arrays according to their sorting

we can choose from anything from any place

its not sorted subarrays

its jsut 5 sub arrays

[fiery cosmos]

yeah I am not choosing sorted subarrays

I am just choosing 4 points to split by brute force

[raven kite]

oh i got it

brute force is possible

[fiery cosmos]

also this explanation seems to be wrong

[raven kite]

why

"split" word mb

[fiery cosmos]

you need sub-array not sub-sequences

subarrays are like continous portions

i...j all included

[raven kite]

english is not my native language

you are right

i need 5 partiton of array

[fiery cosmos]

okay a sub array is like there are no break between start and end

[1, 2, 3, 4]

[2, 3] is a subarray

[2, 4] is not

[raven kite]

not 1,4

hm

[fiery cosmos]

yes

[raven kite]

sorry for it

[fiery cosmos]

in your example the grouping doesnt seem correct

it doesn't obey subarrays

so is it sub-array or sub-sequences?

cause the whole algorithm gets changed for both

[raven kite]

sub sequences

think like its group

[fiery cosmos]

reeeeeeeee that is hard

[raven kite]

yea

brute force is mb only solution

[fiery cosmos]

brute force won't pass

it would be like 5^n solution lol

does this element belong to first part or second part or so on ...

[raven kite]

nope

if there is a big element on the last

it will fail

1 ,2 ,3, 4 ,5, 15 for example

[fiery cosmos]

the only thing I can think of right now is recursive + memoization

[raven kite]

it will be minimum 16

but i could be 15

yeap

thanks anyway

[fiery cosmos]

and I am scared to write recursion in python

also how large can an element in the array be?

[raven kite]

no limit

i dont need to test code btw

i need just implement it

after pseudocode

but only brute force is coming to my idea

[fiery cosmos]

okay I might try to code it

I feel like you could you some backtracking here

backtracking will be too slow

yeah yeah

I thought of a binary search solution

ig that should work

O(N*log(sum(array))

but if N is like 10^9 then rip

For binary search minimum sum can be 1 and the maximum sum can be the sum of all the elements.

Maintain a count of sub – arrays, include all possible elements in sub array until their sum is less than the max sum. After, if the count is less than or equal to 5, then sum is achievable .

[ancient turtle]

kill me

[fiery cosmos]

sorry, im bad at explaining stuff 😦

@raven kite for this [10,10,20,30,30,10,20,15,10]

my code found 30

[30] [30] [10,10,10] [15] [20,10]

[raven kite]

there is two 20

[fiery cosmos]

rip wrong input

[raven kite]

its still good soluiton

but [10,10,10] [15] must be [10,10] [10,15]

[fiery cosmos]

no it is giving wrong for [1, 2, 3, 4, 5, 15]

[raven kite]

for better solutions

[fiery cosmos]

I am really worried about the max number in the array

[fiery cosmos]

import numpy as np

n = int(input("matrix dimension: "))

matrice = np.zeros(n, n)

matrix dimension: 3
Traceback (most recent call last):
  File "myfile.py", line 7, in <module>
    matrice = np.zeros(n, n)
TypeError: Cannot interpret '3' as a data type```

why tho?

the second argument takes data type

what?

import numpy as np
d = np.zeros((2,3))

>>> d
    [[ 0.  0.  0.]
     [ 0.  0.  0.]]

matrix = np.zeros((n,n), dtype=int)

should work

yeah

oh, i did it wrong

sorry

[raven kite]

(n^2)! > (n!)^n

how to proof this mathetical

[wraith valve]

i think if u take log of both sides to expand the factorial @raven kite

u will find that the left side has the same terms as the right side plus extra terms

[raven kite]

left side is bigger btw

[wraith valve]

yeah

[raven kite]

log(n^2)! for example

what change

[wraith valve]

left side = right side + extra terms

[raven kite]

right is nlog(n!)

[wraith valve]

n! = n * n-1 etc

log rules

changes to addition

[raven kite]

how to proof n log n bigger than logn^2+....logn

its obvious i see

[wraith valve]

left side expands to log(n^2) + log(n^2-1) etc

[raven kite]

there is at least n time

hmm

its right

but i feel like im adding some comments to it

but thanks

probably its only way

[wraith valve]

@raven kite is there like a range of n that it is valid for

cuz if n = 0

both sides equal 1

and ur inequality isnt true

[raven kite]

its growth rat

so small numbers are not important

mb i can do induction after converting thme log base

@wraith valve

can it be valid?

[wraith valve]

i was thinking the same thing

@raven kite

[raven kite]

so its true @wraith valve

i can do induction

[wraith valve]

i mean prob

but u gotta figure a base case

cuz obviously 0 doesnt work

[raven kite]

3

[wraith valve]

prob

[raven kite]

thanks @wraith valve

[kindred imp]

aghhh, do class projects just get uglier and ugler as you ammend code to bandaid something else 😄

I am sure this is why version1.0 of software sucks

[eager hamlet]

https://codeforces.com/contest/831/my
ok so rn i have a n^3 sol

which times out

and even when i redid it in java

it still times out

for the same test case so

https://paste.ofcode.org/375UuaMrfrXQA7pbKKZrkqv

any help? (ping plz)

[ashen wharf]

why does this insertionsort not work

def insertionSort(sort):
    for i in range(0, len(sort)):
        if sort[i-1] > sort[i]:
            for j in reversed(range(0, i)):
                if sort[j-1] > sort[j]:
                    sort[j-1], sort[j] = sort[j], sort[j-1]
                else:
                    break

    return sort

[drifting drum]

for i in range(0, len(sort)):
    if sort[i-1] > sort[i]:
``` that part is gonna compare the -1th (last) element with the 1st element of the list(in the first iteration), is that what you really wanna do? Or should`i` start from 1 ?

[ashen wharf]

it starts from 0

[drifting drum]

I know it starts from 0

are you sure you read what I wrote correctly?

[ashen wharf]

it won't compare the last and the 1st element

oh wait

ok i'm dumb but how do i fix it

i literally made a working insertionsort a week ago

i made it start from 1

modified it

def insertionSort(l):
    for i in range(1, len(l)-1):
        if l[i-1] > l[i]:
            for j in reversed(range(0, i+1)):
                if l[j-1] > l[j]:
                    l[j-1], l[j] = l[j], l[j-1]
                else:
                    break

    return l
``` @drifting drum

[drifting drum]

yeah I'm trying to trace the bug in the second loop

[ashen wharf]

thank you so much for finally helping

@drifting drum are you there?

[drifting drum]

yeah

so I slightly modified your code

def insertionSort(sort):
    for i in range(1, len(sort)):
        key = sort[i]
        j = i - 1
        while(j >= 0):
            if sort[j] > key:
                sort[j+1] = sort[j]
            else:
                break
            j -= 1
        sort[j+1] = key

the problem with the earlier one I think is that I don't see you hold a reference to the ith element, ie the key

[ashen wharf]

you mean the i?

[drifting drum]

key = sort[i], yes

[ashen wharf]

well, it actually seems to work right now, except

for some reason it seems to mix up the last few elements

i didn't change the code since

[drifting drum]

I'm fairly certain what I wrote works on any input, including the one in the pic

[ashen wharf]

ok

i was talking about my code

[drifting drum]

I did say you don't have a reference to the ith element

I'm not sure you understand how insertion sorts work

[ashen wharf]

i mean is it required to have a variable to store sort[i]?

[drifting drum]

yeah

[ashen wharf]

actually for the rest of the code where the displaced element moves back, i just swapped the elements

it doesn't need to keep track which one is the key since it either swaps and then continues or breaks

this demo shows the list and then which index has been swapped

it does everything right but doesn't check the last element

i do not know why

@drifting drum oof, sorry for another ping

if i'm totally wrong then i'll use your code

[drifting drum]

I've never actually done it without using the key element, let me think

[ashen wharf]

vector<int> insertionSort(vector<int> list) {
  int temp;
  for(int i = 1; i < list.size(); i++) {
    if(list[i] < list[i-1]) {
      for(int j = i; j > 0; j--) {
        if(list[j] < list[j-1]) {
          temp = list[j];
          list[j] = list[j-1];
          list[j-1] = temp;
        } else {
          break;
        }
      }
    }
  }

  return list;
}

this is my cpp code, idk if you understand or not, but it works well and similarly without a key

[drifting drum]

oh

I get the problem

it's in this line
for j in reversed(range(0, i+1)):

[ashen wharf]

yes?

[drifting drum]

That should really be
for j in reversed(range(1, i+1)):
Since you wanna exclude the 0

could you check if that solves it?

[ashen wharf]

it actually doesn't change anything

i tried it

[drifting drum]

one sec lemme get to my screen

def insertionSort(sort):
    for i in range(0, len(sort)):
         for j in reversed(range(1, i+1)):
                if sort[j-1] > sort[j]:
                    sort[j-1], sort[j] = sort[j], sort[j-1]
                else:
                    break
    return sort

[ashen wharf]

hmm

so all along this was the problem with the start and stops XD

tysm!

[drifting drum]

This has one less check

if you really want to stick to your original code

def insertionSort(sort):
    for i in range(0, len(sort)):
        if sort[i-1] > sort[i]:
            for j in reversed(range(1, i+1)):
                if sort[j-1] > sort[j]:
                    sort[j-1], sort[j] = sort[j], sort[j-1]
                else:
                    break
    return sort

[dire meadow]

I have a YAML file, with different materials, when i read it with pyYMML i get a list of list of dicts, i have problems querying this data, i'm quite new to python, but i managed to get it to work, but i'm sure there is a better way, is this the right place to ask ?

[drifting drum]

Maybe not, a help channel seems better suited for that

[dire meadow]

ok, i kinda figured it out, i needed this to be converted to this:
<class 'list'>
[{'materialTypeID': 34, 'quantity': 566667}, {'materialTypeID': 35, 'quantity': 133333}, {'materialTypeID': 36, 'quantity': 37778}, {'materialTypeID': 37, 'quantity': 10667}, {'materialTypeID': 38, 'quantity': 2889}, {'materialTypeID': 39, 'quantity': 1312}, {'materialTypeID': 40, 'quantity': 412}]

<class 'dict'>
{34: 566667, 35: 133333, 36: 37778, 37: 10667, 38: 2889, 39: 1312, 40: 412}

No way to do it but to write a function i guess.

My problem was, how do i lookup a material...

[drifting drum]

well you could just iterate through the list and update your dictionary, perhaps neatly using some comprehensions

[drifting drum]

not sure what you mean

[dire meadow]

the problem is that i have a list of pairs of dictionaries, i needed a single dictionary.

[drifting drum]

do you want to elaborate?

[dire meadow]

I had i list of:
{keyA:valueA}, {keyB:valueB}
i needed a list of
{valueA:valueB}

I need it bacause i have no clue as how to lookup valueB from valueA in the other format.

[drifting drum]

your_list = [{'materialTypeID': 34, 'quantity': 566667}, {'materialTypeID': 35, 'quantity': 133333}, {'materialTypeID': 36, 'quantity': 37778}, {'materialTypeID': 37, 'quantity': 10667}, {'materialTypeID': 38, 'quantity': 2889}, {'materialTypeID': 39, 'quantity': 1312}, {'materialTypeID': 40, 'quantity': 412}]

your_dict = {item['materialTypeID']:item['quantity'] for item in your_list}

[halcyon plank]

You are not allowed to use that command here. Please use the #bot-commands channel instead.

[drifting drum]

is that what you want?

[dire meadow]

ok, that was it, i have written a function doing just that.

[drifting drum]

yeah it's just a single line of code

comprehensions are nice like that

[dire meadow]

just tested, same result, now is this new dict a copy or a reference to the original data.

I will go read more about comprehensions, before i ask more.

[drifting drum]

it's not a reference

infact your values in the orginal nested dictionary are actual literal integers, not objects

[dire meadow]

how do you format code ?, in the chat i mean.

[drifting drum]

!code

[halcyon plank]

Here's how to format Python code on Discord:

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

These are backticks, not quotes. Check this out if you can't find the backtick key.

[dire meadow]

i guess it's the same as:

materialList = dict()
for material in materials:
    materialList[material.get('materialTypeID')] = material.get('quantity')

[drifting drum]

yes, it's the same

the idea was the write that in a compact form, which was why we used a comprehension
but then again that's upto you

[dire meadow]

Well, i would have to declare a function, just because i'm not that clever, seems smarter to learn.

thanks.

[agile sundial]

wait, how are you gonna write an insertion sort without a key

[dire meadow]

@agile sundial are you asking me ?

[agile sundial]

no

[raven comet]

I’m trying to make a program that requires authorisation of a username to use (usernames have been imported from an external file called usernames.txt but whenever I put something in it says access denied for each line in the file despite that I’m putting in the correct username

[sleek tundra]

Not sure if this is the right channel, but is there any way I can easily iterate through all attributes of an object, and all attributes of all the attributes, and all of the attributes of those etc? I'm trying to convert nested dictionaries into an object with attributes, and there doesn't seem to be an easy way.

[ashen wharf]

def partition(low, high, l):
    i = low - 1
    pivot = (low-high) // 2

    for j in range(low, high):
        if l[j] <= pivot:
            i += 1
            l[i], l[j] = l[j], l[i]

    l[high], l[i+1] = l[i+1], l[high]
    return i+1


def quickSort(l, low=None, high=None):
    if len(l) == 1:
        return l

    if not low and not high:
        low = 0
        high = len(l)-1

    if low < high:
        part_idx = partition(low, high, l)
        l = quickSort(l, low, part_idx-1) + quickSort(l, part_idx+1, high)

    return l```

can anyone help

the quicksort isn't sorting properly

[tawny shore]

@sleek tundra py class attrDict: def __init__(self, mapping): self.__dict__ = mapping
Then loop through the dictionary recursively and replace all dicts (including the top level one) with instances that class ^

That will let you make dictionaries accessible thru attributes

[sleek tundra]

That's the issue I have, I can't loop through the dictionary, since there are nested dictionaries inside it.

I've found a module munch which might be what I need though, I'm looking into it right now.

[ashen wharf]

can anyone help with my quicksort code above?

[tawny shore]

for key, value in mapping.items():
     #loops thru all keys+values ``` @sleek tundra

[sleek tundra]

@tawny shore Would that loop through all keys and values for all attributes of attributes of attributes etc.?

[tawny shore]

Well no, you'd need to run it recursively

Basically use isinstance to check if the value is a dictionary, then run the func on it

I can sketch up what the function would look like roughly if you want @sleek tundra

[ashen wharf]

i'm trying to make this work #algos-and-data-structs message

[sleek tundra]

@tawny shore Thanks for the help! I think I've figured it out.

[fiery cosmos]

num = "53,34.4,56.8,43.8,16.9,9.6,4.7,5.8,23.5,48.2,66.6,97.6"

How do i make it a list without changing anything?

[cloud bough]

can someone recommend me algorithms to learn? like the most common used ones. I am completely new to this, i only know OF linked lists and binary trees

[grizzled schooner]

you can use string.split() to split a string into a list. num.split(“,”) would split your string anywhere there’s a comma

[ashen wharf]

@cloud bough sorting algos

[cloud bough]

https://i.imgur.com/ZSiURTs.png

such as these?

[ashen wharf]

yes

"selection sort, bubble sort, insertion sort, merge, quick sort, heap sort" are pretty much the 1st you should learn

[cloud bough]

awesome, ill get to learning then. thanks

[fiery cosmos]

Thanks.

1 last question.

how do i add these to a list?

will append work?

[grizzled schooner]

could you show your code?

[fiery cosmos]

import json

with open("climate.json", 'r') as f:
    data = json.load(f)
bis = []
sum = 0
total = 0
n = -1
hello = ""
cello = ""
climate = ["high", "low", "dryDays", "snowDays", "rainfall"]
city = input("Enter the city")
if city == "0"
for id in range(0, 105):
    if data[id]["city"] == city:
        city_copy = data[id]["monthlyAvg"]
        while n <= 10:
            n += 1
            for record in climate:
                nis = data[id]["monthlyAvg"][n][record]
                hello = nis + sum
            h = bis.append(hello)
        print(bis)
        break

its basically extracting data from a json file with nested dictionaries

i got it to work thanks

[lunar jacinth]

Does anyone have some experience in huffman greedy algorithms?

i'm trying to implement this but im kinda stuck

[lunar jacinth]

import heapq
import collections


def huff(my_freq_data):
    merged_lists  = []
    heaps = []

    for i, my_lst in enumerate(my_freq_data):
        if my_lst:
            heaps.append((my_lst[0], i, 0))
    heapify(heaps)

im trying to create something like this

can someone lmk if im on the right track?

does that have to be a dictionary or is a list fine?

its called the huffman greedy algorithm

[bronze sail]

what would be faster? 2d graph with connections or space partitioning?

[clever solstice]

check all value in dict, if match, return the key of the value
have anyway to do that?

[agile sundial]

have you heard of for loops?

[clever solstice]

Yep

but i don't know how

I remember I had seen but I forgot how to do

[hollow isle]

c=0
for i in dict.values():
    if i==match:
       print(dict.keys()[c])
       break
    c+=1

is one possible solution

[clever solstice]

I will try

[agile sundial]

how does that even work

what is c for

@hollow isle

[clever solstice]

i==match?

[hollow isle]

counter..

[agile sundial]

sure, but why

[hollow isle]

print(dict.keys()[c])

[agile sundial]

also, how does dict.keys()[c] work

[hollow isle]

isnt dict.keys() a list

so the c'th element of dict.keys()

[agile sundial]

nvm, i cannot read

[hollow isle]

LOL

[agile sundial]

you can also use dict.items()

[hollow isle]

there are many ways, yes.

[hollow isle]

where match is the matching word you wanna match with

[clever solstice]

the values of dict

[hollow isle]

wha-

i didnt understand

wdym

[clever solstice]

if it match to the values of dict, return the key of value

[hollow isle]

if what match

what is 'it'

[clever solstice]

I'll use it in @commands.command()

async def a(self, ctx, item)

item is 'it'

[hollow isle]

for i in dict.values():
    if i==item:
       print(dict.keys()[c])
       break
    c+=1```

must work

[clever solstice]

@hollow isle await ctx.send add after c+=1 or for?I mean indent

[hollow isle]

c=0
for i in dict.values():
    if i==item:
       await ctx.send(str(dict.keys()[c]))
       break
    c+=1```

[agile sundial]

you can skip on the count variable and the indexing if you use dict.items

[hollow isle]

    if i[0]==item:
       await ctx.send(str(i[1]))
       break```

[clever solstice]

TypeError: descriptor 'items' for 'dict' objects doesn't apply to a 'list' object

@hollow isle

[hollow isle]

bro replace dict with your dictionary name

and dont name your dictionary dict never use keywords for naming

[clever solstice]

why It become list?

[{'a': 'bc', 'c': 'h', 't': 'u'}]

it is list?

When I want py dict(thatlist)
It sendValueError: dictionary update sequence element #0 has length 13; 2 is required

@hollow isle

[hollow isle]

you cant make a list a dictionary

whats the name of list

[clever solstice]

lists = dict(jdata['loadaliases'])

@hollow isle

[hollow isle]

i dont understand lol

guys i need help

i wanna convert a dictionary

to markdown table

here's the dict i have

{'Airline Rank': 'Captain', 'Base Airport': 'OMDB', 'IVAO VID': '43', 'VATSIM CID': 'Not Provided', 'Pilot Origin ': '![United Arab Emirates] (AE)', 'Total Pay ': '138,285 /-', 'Status': ' Active '}

can you convert this using python to something to a markdown table

i saw the tomark library but that's makes the keys as headings and values as values, however i want it like
+--------------------------+
|Airline Rank | Captain |
| Base Airport| OMDB |

like that

can someone help

[clever solstice]

also I import a dict from json, and it become list, how make it conver to dict?

[hollow isle]

learn the json library https://docs.python.org/3/library/json.html

[clever solstice]

my list like this [{'a':'ab', 'b':'bc'<etc>}]

[clever solstice]

wym

[hollow isle]

lol wym wym, you cant

[clever solstice]

"make a list a dict"?

[clever solstice]

wym you cant

[flat sorrel]

If your list only has a single dict inside then you can get the dict using my_list[0]

[clever solstice]

async def load(self, ctx, extension):
    lists = jdata['loadaliases'][0]
    for i in lists.items():
        if i[0] == extension:
            b = i[1]
            self.bot.load_extension(f'cmds.{extension}')
            await ctx.send(f'load **{extension}**')
            break```
Why it don't send anything?

[clever solstice]

    async def load(self, ctx, extension):
        lists = jdata['loadaliases'][0]
        for i in lists.items():
            print(1)
            if i[0] == extension:
                print(2)
                b = i[1]
                for a in lists.keys():
                    print(3)
                    if a == extension:
                        print(4)
                        self.bot.load_extension(f'cmds.{extension}')
                        print('a')
                        await ctx.send(f'load **{extension}**')
                    else:
                        self.bot.load_extension(f'cmds.{b}')
                        print('ab')
                        await ctx.send(f'load **{b}**')
                break```Why it doesn't work? @hollow isle

[sweet cipher]

Walrus operator variable leaking is fine for global use, but even when the walrus operator variable in the list comprehensions also leaking to globals is acceptable?

Should we not report it as a bug for it

>>> [(t:=(t[1], sum(t)) if i else (0,1))[0] for i in range(20)]
[0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181]
>>> t
(4181, 6765)

[halcyon plank]

You are not allowed to use that command here. Please use the #bot-commands channel instead.

[sweet cipher]

In the above example the t variable is leaking into the global-namespace

Even when in the list-comprehension

[fast birch]

Hello ! I have a huge XML file with the following structure:

<?xml version="1.0" encoding="UTF-8"?>

<!-- (...definition of elements and entities...) -->

<JMdict>
<entry>
<ent_seq>1000000</ent_seq>
<r_ele>
<reb>ヽ</reb>
</r_ele>
<sense>
<pos>&unc;</pos>
<xref>一の字点</xref>
<gloss g_type="expl">repetition mark in katakana</gloss>
</sense>
<sense>
<gloss xml:lang="dut">hitotsuten 一つ点: teken dat herhaling van het voorafgaande katakana-schriftteken aangeeft</gloss>
</sense>
</entry>

<!-- (...5 million lines...)  -->

</JMdict>

It's a multilingual Japanese dictionary. It's 104Mo. How can I perform binary search in the dictionary's entries without loading all of it in memory?

[trim fiber]

@fast birch There is some tool called duplicut https://github.com/nil0x42/duplicut so maybe it is going to be helpful

[fiery cosmos]

How do i get the minimum value from a dictionary?

[agile sundial]

min

[fiery cosmos]

i wanna get the minimum value from the monthly dict.

will min work on " monthlyAvg"??

min(for value in dict["montlyAvg"].values())

does this work?

[cloud bough]

does someone have good courses for DS and Algos?

free preferred 😄

[tough ginkgo]

Hey guys I have a task about Python Alg. This task is from my school. I was tried so many ways to solve this problem but I couldn't make it. That task so important for me. Could anyone help me pls?

[fiery cosmos]

is this homework?

@tough ginkgo

[tough ginkgo]

unfortunately it is and this is final task from my homework. I have to handle it.

[fiery cosmos]

maybe I can give you a hint
|| count the number of distinct number of characters using ___ and then subtract this from the total length. Think why? ||

[tough ginkgo]

it can be help but I wanna ask something else. Can you give me more a little bit hint 😦

[fiery cosmos]

what part does it seem hard?

[agile sundial]

have you thought about the previous one?

[tough ginkgo]

I was working on 3 task before fourth. This tasks tired me of. Now I can't even think. Normally, it seems not hard to me. But unfortunately I can't anything about that because I am so tired now and I have to deliver tomorrow morning. For that reason I written here to get help. Sorry guys I did not want to bother anyone.

[fiery cosmos]

guys what is this window called?

a text box inside a canvas?

idk I am not good at gui stuff

well my prof didnt teach us gui. so im kinda lost lol

i could maybe do some research if i knew what it is

[fiery cosmos]

nvm. my prof was using another ide, no wonder i didnt get it lol

[fiery cosmos]

what is the use-case for algorithms like merge-sort when you can just use something like sort() in python?

[agile sundial]

well...sort uses a hybrid mergesort

so there's that

[ashen wharf]

def sort(l):
    res = [min(l), max(l)]
    l.remove(min(l))
    l.remove(max(l))

    for i in l:
        for interval in range(len(res)-1):
            if res[interval+1] > i > res[interval]:
                res.insert(interval+1, i)
                break

    return res
```does this algorithm match any existing one?

[mint jewel]

looks like insertion sort

[ashen wharf]

yeah it's similar

the only thing is that i'm not iterating in the result list, as doing that along with using insert might be a problem...

[zenith pagoda]

Why it is said that for a binary tree to be balance is that the difference between the left and right subtrees should not be more than 1? Why 1? Is there any meaning behing the number 1?

[agile sundial]

it means the difference must be 0

[zenith pagoda]

i'm sorry but could you please elaborate it?

let me share this code from CTCI

public static int getHeight(TreeNode root) {
  if (root == null) {
    return -1;
  }
  return Math.max(getHeight(root.left), getHeight(root.right)) + 1;
}
        
public static boolean isBalanced(TreeNode root) {
  if (root == null) {
    return true;
   }
  int heightDiff = getHeight(root.left) - getHeight(root.right);
  if (Math.abs(heightDiff) > 1) {
    return false;
  }
  else {
    return isBalanced(root.left) && isBalanced(root.right);
  }

[agile sundial]

that doesn't look like python

[zenith pagoda]

oh sorry

let me change it

[wraith valve]

i think what ur saying is the definition of avl tree

[zenith pagoda]

@wraith valve I'm not sure. I'm not yet there on that topic

[wraith valve]

like the difference in height between 2 subtrees cannot be greater than 1

thats just definition

like u can also choose another definition which is what red black uses

[wide prism]

not sure, but i guess allowing the gap to be one allows some flexibility as to the number of nodes that can be in the tree
eg if you required the depths to be the exactly the same you could not balance a tree with two nodes, or with 4 if you wanted to fill every level other than the last
maybe someone can say with more authority

btw wikipedia has

A balanced binary tree is a binary tree structure in which the left and right subtrees of every node differ in height by no more than 1.[22] One may also consider binary trees where no leaf is much farther away from the root than any other leaf. (Different balancing schemes allow different definitions of "much farther".[23])
so i'm thinking the right way to go is a bit dependent on the context at hand

[wraith valve]

i think the difference of no more than 1 asserts the log n runtime

not sure about a different size tho like 2 or 3

[wide prism]

yeah, i took the question to be "why not 0? why not 2?" etc

[zenith pagoda]

@agile sundial
this is the python equivalent

def get_height(root):
    if root is None:
        return -1

    return max(get_height(root.left), get_height(root.right)) + 1

def is_balanced(root):
    if root is None:
        return True

    height_diff = get_height(root.left) - get_height(root.right)

    if abs(height_diff) > 1:
        return False
    else:
        return is_balanced(root.left) and is_balanced(root.right)

[lean glade]

Does this array have the form of a binary heap? Just to trying to understand what Im doing. [4, 6, 7, 3, 5, 2, 1]

[wraith valve]

i think simple is doing balanced trees

oh wait ur not talking about simple

@lean glade assuming ur doing a min heap the 3 5 children of 6 doesnt seem to work

[zenith pagoda]

I think you're right about this

[lean glade]

I think it's a max heap?

I don't know much, if not nothing

[wraith valve]

if it was a max heap 4 is in the wrong position

[zenith pagoda]

thank you for citing this

[wraith valve]

honestly idk what will happen if the difference in height was like at max 2 or 3 instead of 1

[lean glade]

So should I do binary heaps out of lists by sifting down or by sifting up?

[wraith valve]

uh for a better runtime u want to heapify down

starting from halfway

and go to the start

[lean glade]

But what I don't understand is that from what Im reading here (yes, in wikipedia), sifting down is for extraction?

[wraith valve]

and with regard to difference in heights i think u just need a limit on how much the difference can be. you cant have too big of a difference so i guess they just chose 1

@lean glade so if u want to remove for a priority queue

u pop the first

and u replace that with the last element

and u heap down the first element

[lean glade]

And where do I place that popped element?

[wraith valve]

oh that is gone

u removed it

like its not literally gone

[lean glade]

But when Im building the heap

[wraith valve]

but like its not part of the structure anymore

if u want to build a heap from a list

u start from index n/2 and u go to index 0 assuming u have an 0 index heap

and for each element u heapify down for the better runtime

[lean glade]

Hmm ok

I'll see what I can do

Thx

[zenith pagoda]

@wraith valve

May I ask why this is not a balanced tree?

      10
     /  \
   20    40
  /     /   
30    50     
     /
   60 

[shrewd cradle]

I don't think this is height-balanced. The right branch can be more optimized

[wide prism]

oops i'm taking that back

A balanced binary tree is a binary tree structure in which the left and right subtrees of every node differ in height by no more than 1

eg the node labeled 40 has a subtree of height 2 and and one of height 0, so

(sorry, i misread the def before making the comment preceding this one)

[zenith pagoda]

is it the same to what @wide prism is saying?

[zenith pagoda]

oh ok I think I understand now, so let's say if there's another node on the right of 40 then it will be balance? right? because that will make a difference of 1 against the left side of 40 ?

[shrewd cradle]

Yeah I didn't provide the details but vt explained it well

[zenith pagoda]

yeah but still thank you

[wraith valve]

u need to do a rotation with the tree that has 40 as a root

[zenith pagoda]

@wraith valve no I mean if I will create a new tree with the same with that but with a node on the right of 40 will that be a balanced tree?

[wraith valve]

if u insert sth to the right of 40 it should

cuz 40's left subtree has a height of 1

and 40's right subtree has a height of 0

and then u check 10 who's left subtree has height of 1 and whos right subtree has a height of 2

[zenith pagoda]

@wraith valve Thanks,

last question, may I ask why in this code why they add 1 after getting the max between the root.left and root.right ?

 return max(get_height(root.left), get_height(root.right)) + 1

[lunar jacinth]

does anyone know why my program isnt working correctly?

import heapq
import collections

def helper_func(k):
    return len(k[-1]), k


def huff(my_freq_data):
    heaps = []
    for sym, freq in my_freq_data.items():
        heaps.append([freq, [sym, '']])
    heapq.heapify(heaps)

    while len(heaps) > 1:
        l = heapq.heappop(heaps)
        h = heapq.heappop(heaps)
        for i, k in enumerate(l):
            if i != 0:
                k[1] = f"0{k[1]}"
        for i, k in enumerate(h):
            if i != 0:
                k[1] = f"1{k[1]}"
        heapq.heappush(heaps, [l[0] + h[0]] + l[1:] + h[1:])
        return sorted(heapq.heappop(heaps)[1:], key=helper_func, reverse=False)


if __name__ == '__main__':
    string = "this is an example for huffman encoding"
    my_freq_data = collections.defaultdict(int)
    for sym in string:
        my_freq_data[sym] = my_freq_data[sym] + 1

    dummy = huff(my_freq_data)
    print("Symbol".ljust(10) + "Weight".ljust(10) + "Huffman Code")
    for k in dummy:
        print(k[0].ljust(10) + str(my_freq_data[k[0]]).ljust(10) + k[1])

[wraith valve]

ugh

ur tree grows so u gotta add 1

[lunar jacinth]

its only printing out 'g' for symbol, and 1 for frequency

[wraith valve]

like if u take the node 50

its left has height 0 and the right has a height -1 assuming no node is represented as -1 height

but the node itself is one taller than it's tallest subtree so u add 1 to get the node's height

which in the case of 50 is 0 + 1 = 1 which matches the pic

[zenith pagoda]

@wraith valve I'm sorry is that you answer here? or still about adding a node on the right of 40 ?

[wraith valve]

we just looking at node 50

so it has one child to the left and none to the right

[zenith pagoda]

sorry for I'm confuse that's for my new question, isn't?

[wraith valve]

thats for the min(left, right) + 1 = height

[zenith pagoda]

oh ok but why min()? I think it should be max()

[wraith valve]

ya u right it is max

typo

[zenith pagoda]

ok so with this diagram isn't it when we take the node 50 the height on the left will be 1 and right is 0?

[zenith pagoda]

but you said here the height will be 0 and the right will be -1

[wraith valve]

ur counting the heights of the subtrees

not the tree itself

[zenith pagoda]

I'm sorry I'm lost, may you elaborate that part?

[wraith valve]

like when u find the height of the tree

u find the maximum height of the subtrees

and to find the height of the subtrees u ignore the node itself and look at the child nodes

and u count the height of the child nodes

[zenith pagoda]

let me cut you on that part

on the diagram, you mean I will ignore the 10 and look on the child of it on both left and right?

[wraith valve]

yeah

[zenith pagoda]

ok continue

[wraith valve]

u find the height of the tree that has root 20

and u find the height of the tree that has root 40

those two are subtree heights

and now to find the height of the tree that has root 10

u find the max of the height of the tree that has root 20 and the tree that has root 40

and then u add 1

[zenith pagoda]

cut

ok for example

for the tree with the root 20, what will be its height?

[wraith valve]

the height is 1

[lean glade]

Is this now a maxheap? [7, 3, 5, 6, 4, 2, 1]

[wraith valve]

uh no the 3 is in wrong position

[lean glade]

Should it be after the 5?

[zenith pagoda]

that height consider both left and right of 20, isn't?

[wraith valve]

yeah

the height of the tree left of 20 is 0 and the height of the tree right of 20 is -1 bc there is no tree

[zenith pagoda]

wait why 0? not 1?

[wraith valve]

what is the height of one node

[zenith pagoda]

you mean a tree with one node?

[wraith valve]

yeah

if the tree has one node how tall is it

[zenith pagoda]

what I understand is that it is 0 because if that tree has a child that will make it's height 1

I'm not sure

[wraith valve]

yup if a tree has a child it is height one

so the height of the subtree 30 is 0

so the height of the subtree 20 is 1

[zenith pagoda]

oh wait

[wraith valve]

there is a distinction between the height of the current tree and the height of the subtrees of the current tree

[zenith pagoda]

I'm lost with you wording here

[wraith valve]

the tree left of 20 is the tree 30

[zenith pagoda]

yeah

[wraith valve]

the tree 30 has height 0

the tree right of 20 is no tree

[zenith pagoda]

ok I understand now

[wraith valve]

no tree has height -1 in this case

[zenith pagoda]

ok

wait

[wraith valve]

so to find the height of tree 20 u first need to know the height of tree 30 and no tree

[zenith pagoda]

you said here

is the same with no tree

[wraith valve]

yeah 30 has height 0

no tree has height -1

[zenith pagoda]

but why you said it is -1 in this case

[wraith valve]

30 is a tree

but the thing to the right of 20 is not a tree

its a NULL

[zenith pagoda]

oh ok

continue

[wraith valve]

so since 30 has height 0 and the NULL has height -1

the height of 20 would be the max of those two + 1

so max(0, -1) = 0

0 + 1 = 1

so height of 20 is 1

[zenith pagoda]

ok

so this is the part i'm confuse

what is +1 for?

why do we need to account it?

[wraith valve]

do u know the difference between a tree and a subtree

[zenith pagoda]

no, what do you mean?

I'm sorry

yeah of course

[wraith valve]

for the tree with a root 20

what is one of the subtrees

[zenith pagoda]

what is one or which one? if which one then the tree with the root 30

@wraith valve

[wraith valve]

i mean the tree with root 20 in ur example

a bin tree has two subtrees

what is one of the subtrees of the tree with root 20 in ur case

[zenith pagoda]

I'm confuse with your wording here I'm sorry

@wraith valve

[wraith valve]

if u have a tree

   20
  / 
30  

what are the subtrees of it

[zenith pagoda]

@wraith valve the tree with root 30 and a null tree

[wraith valve]

okay good

so what is the height of the tree with root 30

[zenith pagoda]

0

[wraith valve]

good

so to find the height of the tree 20

u need to know the height of the tree 30

since 20 has one more edge to it, it's height is 1 greater than the height of 30

which is where the +1 comes from

[zenith pagoda]

oh

wait

I quite understand it now but I need to a proper wording to remember it

so +1 is needed to account for the height of current node as being parent of two subtrees?

[wraith valve]

yup

[zenith pagoda]

which is the same when we need to find the height of tree with root 10 (the entire tree). We neglect the node 10 itself for the moment but rather focus on its subtrees (on left and right)?

is that right?

[wraith valve]

yeah

once u found the subtrees

u can then use the max value of the subtrees then u add 1 to account for the extra edge

[zenith pagoda]

awesome

[main hedge]

How can I extract all "metrics" from a JSON that looks like:

{ comments: [
comment: "blahblah"
metric: "metrics"
replies: [
  reply1: {
      comment: "blahblah"
      metric: "metrics"
      replies: [
        reply1: {
          comment: "blahblah"
          metric: "metrics"}
        reply2: {
          comment: "blahblah"
          metric: "metrics"}
       ]
      }
  reply2: {
    comment: "blahblah"
    metric: "metrics"
    replies: []
    }
  ]}

I have no idea how deep the rabbit hole goes (meaning I have no way of knowing if there is one reply to a given comment or 20). Right now I've been saying:

try:
  for comment in JSONofabitch['comments']['comment']:
      getMetrics(comment['metrics'])
      try:
         for reply in comment['replies']
            getMetrics(reply['metrics'])
        ...
       Except as KeyError:
          continue
Except as KeyError:
  continue

[eager hamlet]

https://www.codewars.com/kata/5b84d6d6b2f82f34d00000d7/discuss

so uh

how the frick do i do this

i'm guessing there's maybe an O(1) way to do each of these?

i mean an O(1) way to apply an operation on a row?

[runic tinsel]

like you use O(n) time to make O(n) steps

so that's kinda O(1) per row

idk it's confusing

I did the "insane" one, technically a different kata

yeah, can't pass this one

[brave oak]

How can I extract all "metrics" from a JSON that looks like:

{ comments: [
comment: "blahblah"
metric: "metrics"
replies: [
  reply1: {
      comment: "blahblah"
      metric: "metrics"
      replies: [
        reply1: {
          comment: "blahblah"
          metric: "metrics"}
        reply2: {
          comment: "blahblah"
          metric: "metrics"}
       ]
      }
  reply2: {
    comment: "blahblah"
    metric: "metrics"
    replies: []
    }
  ]}

I have no idea how deep the rabbit hole goes (meaning I have no way of knowing if there is one reply to a given comment or 20). Right now I've been saying:

try:
  for comment in JSONofabitch['comments']['comment']:
      getMetrics(comment['metrics'])
      try:
         for reply in comment['replies']
            getMetrics(reply['metrics'])
        ...
       Except as KeyError:
          continue
Except as KeyError:
  continue

@main hedge recursion would be good

[severe spire]

Any book recommendations for algorithms and data structure?

[viral lily]

Data Structures and Algorithms Made Easy by Narasimha Karumanchi

[vivid horizon]

Introduction to Algorithms

@severe spire It's probably the best book on the subject.

[fiery cosmos]

alr can someone help

1. A linear search would go one by one finding out which one is the same as the value your give it.

2. While the binary finds it in one go.

is it right

[keen hearth]

You're right about linear search, but a binary search works by repeatedly ruling out half of the remaining options.

Is this a graded test?

[main hedge]

Honestly, I forgot that python functions can call themselves. Thanks for the tip.

[pure gust]

not sure if this is the right channel, but I'm basically looking for some help with something I'm trying to do with game screenshots.
Want to determine the native resolution of a game, even though its streamed at a higher resolution, if that makes sense.
talking about cloud gaming services here.
those work in a way where the game is 1080p for example, but streamed at 4k (no real upscaling in the game, the stream is upscaled)
I have found a way to do this by hand, but its pretty tedious...
so I'm trying to automate it.
right now I've used PIL to do this, but that just reads the resolutions of the file, which isn't what I'm looking for.
anyone an idea how I could do this?

[fiery cosmos]

i have a text file. and inside it there’s like
test1
test2
test3
if i look for a string in the text file it finds it. but say i look for t. it also finds it and lets it through. is there a way i can match case
so it only accepts the ones that are in there
such as test3 and not just 3

[royal holly]

Leetcode > algoexpert

[zenith zodiac]

I want to do a list of n numbers that the user types, I want the result as another list which has the same numbers from lowest to highest, only using for (comparing i and i+1 as numbers), while, and functions.
For example, I type n as 5, it will request me 5 numbers, I type 6, 8, 2, 5, 1
I get a result of a list which has numbers as following: 1 2 5 6 8

[lean glade]

Does the left child need to be bigger than the right child in a max binary heap?

[fiery cosmos]

Hello folks. I need to find the coordinates of the intersections of one segment with several others (<10) on the same line. It looks like "interval trees" could be of use, does anybody have experience with them? Could you point out some decent resources to learn about their implementation?

[small sage]

@smoky lion

[smoky lion]

cant zoom in far enough on my phone

[small sage]

# Definition for a binary tree node.
# class TreeNode:
#     def __init__(self, x):
#         self.val = x
#         self.left = None
#         self.right = None

class Solution:
    def getTargetCopy(self, original: TreeNode, cloned: TreeNode, target: TreeNode) -> TreeNode:
        if not cloned or original:
            return
        if original == target: return cloned
        self.getTargetCopy(original.left,cloned.left,target)
        self.getTargetCopy(original.right,cloned.right,target)

code ^

question:

Given two binary trees original and cloned and given a reference to a node target in the original tree.

The cloned tree is a copy of the original tree.

Return a reference to the same node in the cloned tree.

Note that you are not allowed to change any of the two trees or the target node and the answer must be a reference to a node in the cloned tree.

Follow up: Solve the problem if repeated values on the tree are allowed.

example:

Input: tree = [7,4,3,null,null,6,19], target = 3
Output: 3
Explanation: In all examples the original and cloned trees are shown. The target node is a green node from the original tree. The answer is the yellow node from the cloned tree.

[keen hearth]

Nope, only the relative size between parent and child matter in a heap, not between siblings.

[smoky lion]

i know nothing about any kind of tree

[small sage]

well i think @keen hearth does

[lean glade]

Thanks!

[keen hearth]

No prob 👍

[small sage]

anyone good with algos here? i need some help with my Q

[oblique panther]

@small sage go ahead and ask what you would ask if someone said yes.

[keen hearth]

Hey @small sage I suppose your only option is to search for the target node, and remember the steps required to get to that node from the root.

Then follow the same steps in the cloned tree to find the right node.

If the nodes had a parent attribute it would be much easier 😄

[lean glade]

I've been trying to understand how to make a list into a heap, and I found that sifting down through the list is the best option. And from what I understood, it said that I don't need to sift down the entire list, instead I can sift down all nodes which have children, in reverse order, and the last node with children can be found by doing (len(list)//2)-1. Is what I said correct?

[keen hearth]

Yes that sounds right @lean glade

Might need to check the calculation, one sec...

Yeah so, for each two nodes you add to the tree, you make it so that one more node has a child.

For a tree with zero or one nodes there are zero nodes with children. For a two node there is one node with a child. So for an n node tree, there are n//2 nodes with children, the last of which has index n//2 - 1.

Proof by induction 😄

And yes, if you use sift down nodes in reverse order I believe the heapification is O(n), where n is the number of nodes.

[small sage]

@oblique panther

@keen hearth yep thats what i did actually

here is my code:

class Solution:
    def getTargetCopy(self, original: TreeNode, cloned: TreeNode, target: TreeNode) -> TreeNode:
        if not cloned or original:
            return
        if original == target: return cloned
        self.getTargetCopy(original.left,cloned.left,target)
        self.getTargetCopy(original.right,cloned.right,target)

i fuckign got it

i forgot the keyword RETURN

LMAO

oh and also

shouldve said return this or this

not just return line one and return line two

thanks guys

[lean glade]

I wonder why is that

And for the sifting down, I just swap the current node with the root node for each node? Or do I insert it at the top?

[lean glade]

I got to do this using down sifts, but there's something missing, and I really don't know what:

def heapify_down(lis):
    last_parent=len(lis)//2-1
    for ind,t in enumerate(list(reversed(lis[:last_parent+1]))):
        ind=last_parent-ind
        while (2*ind+1 and 2*ind+2)<=len(lis) and lis[ind]<(lis[2*ind+1] or lis[2*ind+2]):
            if lis[ind]<lis[2*ind+1]:
                lis[ind],lis[2*ind+1]=lis[2*ind+1],lis[ind]
                ind=2*ind+1
            elif lis[ind]<lis[2*ind+2]:
                lis[ind],lis[2*ind+2]=lis[2*ind+2],lis[ind]
                ind=2*ind+2

    print(lis)```

[wraith valve]

i think it might be easier to do it recursively

it might also help if u make a function that returns the left child and another function that returns the right child

@lean glade

[lean glade]

What would be the use of a right and left function?

If I can do it with whatever of both children

[wraith valve]

use those as helper functions

basically returns index of left and right child

and u basically then compare

[lean glade]

But that's just 2i+1 and 2i+2

[wraith valve]

sure

is ur thing 1 indexed or 0 indexed

[lean glade]

0 indexed

[wraith valve]

okay

well it might be sth to do with ur while loop

hence y i said if u made it recursive it might be easier

[lean glade]

Yeah, tomorrow I'll implement it recursively

I just wanted to do it with a while loop to help me understand down sifts and all that shit

[wraith valve]

i found it was easier to visualize recursively

[trim fiber]

@karmic quail is it related with topic of the channel?

[karmic quail]

is it? im new to discord and confused

[trim fiber]

I think not 🙂 It rather #tools-and-devops in my opinion

You can always pick some free help channel

#❓｜how-to-get-help

[karmic quail]

alright, i found that this python itself is a channel and it has multiple channels inside it

[sweet cipher]

Microdict on PyPi showing huge gain interms of space,time. I think it's better to implement it in standards track of Python, though the API is somewhat difficult to use. Did anybody seen yet microdict ?

[hoary bronze]

hey how r u guys , i hope u all okay . I want to start algos and data structures. And im completely beginner , so can you guys help me about it (where can i start , how can i start)? can you guys offer me any source ? (video ,book)

or advices , or anything that would help

[blazing crest]

@hybrid ferry

[trim fiber]

Introduction to Algorithms by Cormen, Rivest and others

[hoary bronze]

thx you so much

[keen hearth]

Hey @hoary bronze 🙂 Also check out Khan Academy's algorithms course for a gentler introduction.

I'll find the link....

[autumn timber]

CLRS is outstanding. I’m currently taking DS at university and it is a fantastic supplement. But it is very heavy

[keen hearth]

https://www.khanacademy.org/computing/computer-science/algorithms

CLRS is a really good undergraduate algorithms textbook. But I'd recommend it as your second introduction to algorithms as you might find it easier to digest if you already have some familiarity with the topic.

It's a fairly mathy book.

[autumn timber]

Having a structured series like ^ or Princeton’s free coursera course is the move. Then supplement understanding with books like CLRS or skienna, or videos I.e Abdul bari, backtobackswe, hackerrank

[keen hearth]

Video lectures to accompany CLRS from one of the authors: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-046j-introduction-to-algorithms-sma-5503-fall-2005/

[autumn timber]

Wasn’t aware of that. Great link

[keen hearth]

Yeah, he's a good lecturer too. Super charming guy 😄

Although the videos are lacking in pixels...

[autumn timber]

Nice. Agreed on the math heavy approach though. Im taking discrete and most of the proofs go over my head

[fiery cosmos]

the mit guy got some kind of award

[keen hearth]

Yeah, understanding mathematical proofs is a skill that takes effort to learn. Although it's worthwhile as the techniques make it easier to reason about the correctness of your programs. For example, yesterday in this channel I did an informal proof by induction to check the correctness of an algorithm.

@autumn timber You should check out the Art of Problem Solving book series. They're really meant to be for precocious kids doing maths competitions, but they're actually not bad as a supplement for undergraduates. The Intermediate Counting and Probability is particularly relevant to discrete math.

[autumn timber]

@keen hearth thats great, thats just what I need. I just picked up a copy of how to solve it to read over winter break, I'll add those books. I have one more semester of discrete.

It's the intuition that I'm lacking. In part because I took 10 years off from school/math, and while I did fine in high shool, I didn't particularly care, so I didn't cultivate a "math brain"

[keen hearth]

Oh, the Polya book? That's a classic 😄

Don't worry, I didn't really develop a love of learning and maths until after I left school.

Mostly by watching Khan Academy videos.

[autumn timber]

Yep. Those khan academy folks are great. Used them a lot during the calc progression. I probably shoulda made use this semester too

[lean glade]

Hey @wraith valve, I found out why it wasn't working, it was because I wasn't swapping the node with it's biggest child

And yeah, the sift down method is way more optimal

Im so happy now

[autumn snow]

+1

[wraith valve]

nice @lean glade

[wanton python]

I might be biased as a student, but Algorithms by K Wayne and R Sedgwick is preety good

There's a free coursera course too!

[grave glen]

Sedgwick is a gold standard

[mint jewel]

What's a smart algorithm for find shortest path but you can break a single wall. My idea with bfs where you keep track in the queue whether a wall was already broken on this path or not ended up too slow.

[zenith pagoda]

any tips on visualizing on understanding recursive?

[wanton python]

@mint jewel shortest path but you can skip one edge?

[small sage]

so im doing this question

and im stuck because heres my code:

# Definition for a binary tree node.
# class TreeNode:
#     def __init__(self, val=0, left=None, right=None):
#         self.val = val
#         self.left = left
#         self.right = right
class Solution:
    visited = []
    output = 0
    def deepestLeavesSum(self, root: TreeNode) -> int:
        if root not in self.visited:
            self.visited.append(root)
            if root.left is not None:
                return self.deepestLeavesSum(root.left)
            if root.right is not None:
                return self.deepestLeavesSum(root.right)
        else:
            current = root
            for i in self.visited:
                current = self.visited.pop()
                if current.left is None and current.right is None:
                    self.output = self.output + current.val
        return self.output

the output i get is:

i dont get why because it technically should be recording

can someone help

[fiery cosmos]

okay traverse the tree and if you hit a node store its value in a dictionary where height of the node is the key

now return the sum(dict[max_height])

[odd trellis]

can anyone give me a little help with an exercise? I need to convert an int to a str in order to read each digit an extract one number passed.

if that digit to extract is not in the number then return -1

I have this ugly thing that I tried

[sudden olive]

the find function will probably be helpful

[odd trellis]

But am I thinking this correctly or am I a long shot from what I should be doing

[sudden olive]

that seems fine if you are trying to replace the char with a ":"

[odd trellis]

my problem is that if I print this it prints the string without being modified for some reason

[sudden olive]

ah yeah

bc strings are immutable

[odd trellis]

I could have done this print(stringNumero.replace(extraer, '')) but I lose the return -1 thing and it's not what I need

[sudden olive]

yeah if you want to do it this way

you'll need something like s = s[:n] + ":" + s[n+1:]

[sudden olive]

if you want to do it this way w/ -1:

if c not in s:
  print(-1)
else:
  print(s.replace(c, ":"))

[odd trellis]

ok got it, thanks

One more thing, I don't know if we are going to be allowed to use .replace

Any work around for that?

[agile sundial]

that's a school assignment?

[unborn hawk]

Yea

am stuck on that

[agile sundial]

we can't write it for you

also, this isn't really the right chat, check out #❓｜how-to-get-help

[unborn hawk]

am so sorry

[agile sundial]

np

[jaunty loom]

what is the big O of py class solutions(): def find_duplicate(lst): answer = [] another_list = [] for ind in lst: if ind in another_list: answer.append(ind) another_list.append(ind) return answer?

[agile sundial]

what do you think

[jaunty loom]

O(n)

[agile sundial]

why

[jaunty loom]

cause only one for loop is present

is that a bad assumption?

[agile sundial]

yeah

[jaunty loom]

so how do i calculate it?

[agile sundial]

you have to look at your code

and understand how it works

what's the big O of in on lists?

[jaunty loom]

oh

is in o(n)?

[agile sundial]

yes

[jaunty loom]

so the code should be O(n^2) right?

[agile sundial]

yes

although, if you kept another_list as a set, it would be faster

[jaunty loom]

what would that change?

[agile sundial]

reduce your big O to O(n)

[jaunty loom]

like this py class solutions(): def find_duplicate(lst): store = {} store2 = set() for i in lst: try: if store[i]: store2.add(i) except KeyError: store.__setitem__(i, True) return list(store2)?

[agile sundial]

not like that

[keen hearth]

Hey, btw, you can do i in store to check if it is in the set.

Although store is actually a dict 😄

[jaunty loom]

wait did is in in list 0(n) but in in a set 0(1)?

[agile sundial]

yep

[keen hearth]

Sets and dicts are implemented as hash tables.

[jaunty loom]

so how about this

class solutions():
    def find_duplicate(lst):
        store = set()
        store2 = []
        for i in lst:
            if i in store:
                store2.append(i)
            store.add(i)
        return store2```

[agile sundial]

why not keep store2 as a list

[jaunty loom]

isn't sets a little faster than list?

[agile sundial]

in what way

[jaunty loom]

i guess its useless since im converting it back to a list anyway

thank you @agile sundial, @keen hearth for the help

[fiery cosmos]

excuse me guys, im a university student and i wanna ask all of u python developers. i have a task from my university to make a list. what inside the list is integer that i input. and the input textbox won't stop until i write 0 number. is that possible?. many thx btw

[agile sundial]

yes, that's possible

[lean dome]

A Python <class 'float'> is double precision, not single precision, so you need float64, not float32

[halcyon plank]

Hey @fiery cosmos!

It looks like you tried to attach file type(s) that we do not allow (.pdf). We currently allow the following file types: .3gp, .3g2, .avi, .bmp, .gif, .h264, .jpg, .jpeg, .mkv, .mov, .mp4, .mpeg, .mpg, .png, .tiff, .wmv, .svg, .psd, .ai, .aep, .xcf, .mp3, .wav, .ogg, .webm, .webp, .flac, .afdesign, .m4a, .csv.

Feel free to ask in #community-meta if you think this is a mistake.

[barren terrace]

hello! So i was given this question to solve:

Given that the area of a triangle is N .
Output three sides of the triangle.

Sample input: N = 7
Sample output:
3.5
4
5.315

From what I can see, this is basically just "Heron's Formula", which is √s(s-a)(s-b)(s-c) in reverse, but I'm not exactly sure how to reverse the formula. Considering there are 3 unknown variables ie a, b and c, isn't it impossible to solve?

[fiery cosmos]

@barren terrace They probably want any 3 possible sides, so fix two of them and find the third

[barren terrace]

@fiery cosmos good point. I'll try that now. Tq!

[fiery cosmos]

Hello. I have a question. I've built a web parser to find a bunch of web urls/links. In what format should I save the parsed links? Should I just plain append them to a .txt file and put a comma in between? The goal is to use them in another function and be ressource-friendly. I should mention that all of the parsed links have the exact same length (36) and formatting. Only the last 6 chars differ.

[trim fiber]

You can save one url per one line - it's easy to read this file and recover results

[fiery cosmos]

@trim fiber Thank you. I will look into it

[lean dome]

No, there's not, AFAIK. A Python float object is always double precision. numpy has single precision floats if you need them, I believe

[hot tartan]

is ther a way to run product without it hogging all my memory?

[compact flame]

def swp_cs(out):
    t_out = copy.deepcopy(out)
    for t, c, cswp in zip(t_out, c_s, c_swps):
        for e in t_out[tk]:
            e_s = e['s']
            if e_s:
                e['s'] = e_s.replace(c, cswp)
    return t_out

This works fine, but is there an algorithm I can leverage for this kind of logic? I'm trying to practice my Python algorithms intuition. Maybe something like a transform if? (Please tag me if you reply)

[oak obsidian]

Can you give an english description/function definition @compact flame ?

It is hard to know exactly what this code is doing without any documentation

[deft sinew]

i made a search algorithm but don't know if i can improve the speed
algorithm goes from the database and checks the values of names, every time it matches one in the database the counter is increased by 1 (for that key-value pair) and the item name with the highest counter appears first.

Example
user enters a list of words: jacket, black, lego

data =
[{"name": "Black Card Nvidia", "price":...}, {"name": "Green Jacket Nintendo", "price":...}, {"name": "Black Jacket Lego", "price":...},{...}]

counter for item index 0 is 1 (matching word(s): black),
counter for item index 1 is 1 (matching word(s): jacket),
counter for item index 2 is 3 (matching word(s): black, jacket, lego) thus index 2 appears at top.

EXAMPLE ENDS HERE
EDIT: Did I accidentally reinvented the wheel for a basic search algorithm 🤦‍♂️

[compact flame]

The function takes a dictionary, and returns a transformed copy of it.
The dictionary has the following structure:

{ 
  'A': [ {'a': 'someText1', 's': 'text_to_modify'}, ...],
  'B': [ {'a': 'someText2', 's': 'text_to_modify'}, ...], ...
}

The transformation: Every s key must be accessed and if its value text_to_modify exists, replace a given word in it such as: e['s'] = e['s'].replace(current_old_word, current_new_word).

current_new_word and current_old_word come from known lists of strings which are zip-iterated through along with the dictionary.