Python can return as many values as you want, and I would take advantage of that.
#algos-and-data-structs
opal oriole
fiery cosmos
this sounds correct
opal oriole
You can then have separate file IO code that can take the failure list and write it out.
In general, try to keep IO separate, as its own chunk.
So you can have a nice input -> processing -> output flow in the code.
Like 3 paragraphs / ideas one after another, rather than mixed.
fiery cosmos
right right. ok! let me try
opal oriole
right right. ok! let me try
Note that with this method the user of the data_read function can ignore the return values (the failure list) and not report it. If you want the user of a function to not ignore it then exceptions can be used. They now must not ignore it or it will abort the program with the exception message. So using an exception can be a way to avoid the user making the mistake of ignoring the errors. But sometimes you don't mind if they ignore the errors (soft errors), so it depends.
fiery cosmos
so it's a bit wonky but here is my strategy:
then i'll return both lists for processing in the main function
fiery cosmos
hmm
i also have an interesting problem that was a problem my dad had to solve in engineering school that im curious about
fiery cosmos
opal oriole
Is that a problem? You have to iterate over all of the inputs.
Or do you mean it's generating the wrong valid list with duplicates?
fiery cosmos
Or do you mean it's generating the wrong valid list with duplicates?
This
Because my if condition is within the loop which checks every char
I guess I just have to dedent my append statement but it wasn’t immediately obvious how to do such a thing
opal oriole
I guess I just have to dedent my append statement but it wasn’t immediately obvi...
Usually the way this kind of thing goes is that there is some variable created before the loop (in this case a boolean) with some default value (in this case something like True (all characters are in the language)) and then during the loop that variable is set or conditionally set, and if it's conditionally set then break out early. After that loop do whatever based on that variable.
Examples include checking if a list contains something / meets some boolean condition, computing a summation, product, min, max, or other function that involves a sequence and ends up as one value as the result (stored in the variable created before the loop starts, and used after).
And if it's a boolean and that controls whether you append or skip something in some way, it's a "filter".
fiery cosmos
I don’t want to break out early though, I just want to append a given item of a list to another list if one of the characters of a string is found to be not in the language
Else if an invalid character is not found, append to ‘valid’
opal oriole
Ok, so you have two appends going on.
fiery cosmos
Right but the issue is that the appends are nested as far in as the doubly nested loop which checks each char in a string for each string in my input set
So naturally there are too many appends occurring
opal oriole
The valid append feels like what I just described.
The other appends in that internal loop, but I don't have the code so IDK.
fiery cosmos
Yeah I’ll give an explicit example tomorrow. Thank you for your help
brittle moat
I need help with this algo. It's running infinitely but I Need it to stop once banana is less than 1 b 2ns or 3as
def solution(S):
S = S.lower()
a = ["a", "a", "a"]
b = ["b"]
n = ["n", "n"]
less_than = 0
counter = 0
if S.count("a") < 3 or S.count("b") < 1 or S.count("n") < 2:
return less_than
while not less_than:
for x in range(len(a)):
S.replace("a", "")
for x in range(len(b)):
S.replace("b", "")
for x in range(len(n)):
S.replace("n", "")
counter += 1
return S```
lean walrus
There are a lot of issues in this code
brazen needle
hi how do i make it so a python script will run when a specific url is opened in chrom
chrome
dusky canyon
Hello everyone, can anyone help me to solve this problem?
Any algorithm I write cannot do the processing quickly. My record for 1,200,000 is 5 minutes.
I am also attaching the code file.
https://www.codewars.com/kata/56a1c63f3bc6827e13000006
halcyon plankBOT
Hey @dusky canyon!
agile sundial
you need a flag variable to know if you should append or not. check after the inner loop so you append outside the loop
fiery cosmos
ok i think it just clicked thanks. the flag goes off when an invalid char is detected and then after each inner loop iteration i can check the flag for append or not append status
hmm do i have to abandon my exception i was trying to use to return the invalid char?
can i append the error string as the exception?
agile sundial
because i is an element not an index
limpid smelt
did you print both d and e?
fiery cosmos
i'm not printing them i'm using them to build strings
fiery cosmos
yeah i see what you're saying. according to what i wrote above, one of them doesn't exist. i need to go back and confirm every element of this list is indeed a tuple
limpid smelt
brb lemme switch OS
agile sundial
because now i is an index. show your loop again?
fiery cosmos
so there is a second error string i'm never adding to my invalid list and i don't know why
limpid smelt
Oh I got it, use d[:] to get the strings
or d[0] + d[1]
cuz d is a list now
and e should be 'string'
fiery cosmos
well index i should be looking at each tuple, so not quite
fiery cosmos
its a list compiled from my input of potentially valid strings to process or invalid strings to reject
agile sundial
it's adding 1 when there are 2 in the input
what do you mean by that
fiery cosmos
what do you mean by that
its failing to append the second error string in the input
agile sundial
char will not be the char that is invalid here
agile sundial
its failing to append the second error string in the input
can you show the input, your code, and the output
fiery cosmos
not publically, can i PM you
limpid smelt
can you share like the structure of prevalid?
like [([int, float], string), ([int, float], string)] or something?
agile sundial
not publically, can i PM you
no. do you not have sample data
limpid smelt
it might not be taking a string as true if the last character is in AGCT
is that intentional?
agile sundial
it might not be taking a string as true if the last character is in AGCT
no, it will set the flag if any char is invalid
it just doesn't log the correct char in invalid
can you show a sample input and output
that is representative of the error
limpid smelt
SMELL should be valid and SMALL should be invalid?
agile sundial
SMELL should be valid and SMALL should be invalid?
both are invalid
limpid smelt
but isnt he checking AGCT?
agile sundial
but isnt he checking AGCT?
huh? it's invalid if there are any characters not in "ACTG"
limpid smelt
ah right, my bad
agile sundial
are you parsing it properly? is prevalid what you expect?
fiery cosmos
yes
wait a tick
oh wow, its failing to get the final char of the final string for some reason...
printed prevalid and it's cleaved off
so weird. i missed bc it happens to not successfully get the final char of the final string, and that's where i happened to put an error
let me put it internal to see if it teases out
limpid smelt
is this the correct expected output?
fiery cosmos
is this the correct expected output?
lms
yes but could you please delete this
i don't want personally identifying information
agile sundial
how is that PII lmao
lean walrus
It cant find wrong char in any string. Try to put Q inside string
Deleting messages is bad
limpid smelt
I didnt change your code at all btw
fiery cosmos
i know its weird
agile sundial
that's still not PII though, unless someone has a database of every person's DNA and that is enough to uniquely identify them
fiery cosmos
i think its bc of one of my strip statements
agile sundial
show your parsing code
fiery cosmos
these were randomly generated
limpid smelt
do plagiarism sites check discord as well now? lmao when I was in college they only checked research papers and stuff from google xD
fiery cosmos
🤷♂️
agile sundial
just generate new ones?
fiery cosmos
yeah i'll have to
agile sundial
also, yeah. Google doesn't index discord and I highly doubt anything would
fiery cosmos
extra work
lean walrus
You mean these sequences?
limpid smelt
you could just use something like passgen to generate alphabetical strings randomly?
agile sundial
it's highly likely you need to show more sample data, so just generate some data to submit and some to show
haughty mountain
!e
from random import choice
print(''.join(choice('ATCG') for _ in range(20)))
imagine if that was the same sequence
halcyon plankBOT
@haughty mountain :white_check_mark: Your 3.11 eval job has completed with return code 0.
ACGTACGATCATCGCCGACA
lol
then wouldn't they see the context? are you not allowed to get help or something lol
¯_(ツ)_/¯ ok
show your parsing code?
limpid smelt
Its okay man, we can just let him be. literally no need to ask for info if they arent comfortable with it xD
agile sundial
Its okay man, we can just let him be. literally no need to ask for info if they ...
it's just random data though lol
limpid smelt
if hes not comfortable with it, it doesnt really matter 🙂
thats why I instead asked for the structure of the datatype
fiery cosmos
so the existing issue is that i'm failing to parse every single char, specifically the last char of the last string is being neglected somehow
limpid smelt
try maybe just line[:]?
fiery cosmos
yeah why is it -1?
limpid smelt
idk xD
haughty mountain
just line.strip()...
fiery cosmos
what would fail every string for the same reason though would it not
limpid smelt
also I dont think you need to do line[]?
haughty mountain
[:] is a noop here
limpid smelt
shouldnt line.split() be enough?
yeah like fiery said
line[].split is not needed
cuz line[:] == line
radiant belfry
this line just gets you everything except the last character in the line, you can just use strip() to get rid of \n
lean walrus
i don't want personally identifying information
Why? You care about some random strings, but dont care about your username (which is probably your name). Why random strings are PII, but your name isnt?
fiery cosmos
yep. on my birth cert. thats me
haughty mountain
you could easily just regenerate the random dataset 
limpid smelt
now someone in this server has had their name leaked xD
okay anyways, line.strip().split('=')
lean walrus
okay anyways, line.strip().split('=')
No, you should also do .strip() on each item of result
haughty mountain
okay anyways, line.strip().split('=')
or better, split first and strip each part
varname, seq = map(str.strip, line.split('='))
aah with map
haughty mountain
map or equivalent, yeah
limpid smelt
but idk if they want to handle two different vars or just keep it in a single var
no dont use that
haughty mountain
or unpack in the first loop as well
limpid smelt
so i = (['S2', 'AGAAATTAGAQ'], 'Q') and i[0] = ['S2', 'AGAAATTAGAQ']
yeah could do that as well
haughty mountain
that way you can give sensible names
fiery cosmos
unpack in first loop?
woah. @haughty mountain and @agile sundial were helping me 6 months ago. crazy
can't believe i've been doing this stuff for so long already
limpid smelt
time sure flies when youre doing something you like or are forced to like xD
unpacking inside first loop
fiery cosmos
i'm not able to see how to purpose it for what i'm trying to do
limpid smelt
you want to output a sentence with S<num> and the randString and the invalid char?
or something like that?
fiery cosmos
i'm going to rebuild in the input sequence from the list and then add some statement like this sequence was not processed because of the following character: then print e in my d, e scheme
limpid smelt
since you know there will only be two elements in the i[0]
idk what output format you want though so
also today I realised .strip() only removes the whitespaces from the end and beginning..
I am stupid
.replace() is so much better xD
fiery cosmos
uh oh
limpid smelt
?
fiery cosmos
so just tried a new error, putting a # into one of the strings, and it reported that it was invalid because of the final character (a valid one) and not because of the symbol
wait let me look
probably can figure this out before writing about it on discord 🤦♂️
limpid smelt
maybe just break the for loop as soon as the string is invalid?
or are you storing all the invalid chars?
fiery cosmos
i'm not, i just added an extra variable to store the invalid char and append that one to my invalid list, should have fixed it up let me test
yep
agile sundial
so just tried a new error, putting a `#` into one of the strings, and it reporte...
I said why beforehand
limpid smelt
which invalid_flag state is the correct one if a string is invalid?
limpid smelt
so just tried a new error, putting a `#` into one of the strings, and it reporte...
yeah its because if the string is invalid, it is just storing the last char instead of the actual invalidating char
fiery cosmos
right its fixed up now
fiery cosmos
now i just have to test with a bunch of valid inputs and make sure it's running as expected
limpid smelt
Niceu! Good luck!
fiery cosmos
thanks. i think i've botched the usual runtime with doubly nested loops, but w.e
where n and m are the length of the input strings
limpid smelt
there are two input strings? I thought input string was 'S1= HABHNDJKMD' ?
fiery cosmos
it compares and finds the longest common subsequence between each pair
limpid smelt
true. I still dont get how you would have multiple input strings, but maybe I missed something
fiery cosmos
do you have CLRS text?
let me just get you the pseudocode
this is the core of the program, not including the print algo
as you can see, it only operates on two input strings at a time
but you can build out a python program to run it pairwise on every pair of strings in an input file
limpid smelt
maybe its the difference between our different countries xD no idea what those arrow symbols are xD
but as far as time complexity I got the point you were making
might just be that its 1:30AM
fiery cosmos
The arrows just indicate where to look in the dp table
limpid smelt
is that like a truth table?
fiery cosmos
different from truth table. a dp table stores pre-computed subproblems so that they can be looked up, rather than continuously recomputed
this particular problem is highly amenable to such an approach, because in two sequences of length 500, the number of combinations of possible subsequences to check is greater than the number of atoms in the known universe
agile sundial
is that like a truth table?
it's just a string that points to the right spot when the array is printed a certain way
fiery cosmos
yes, the arrow is just a string which points somewhere
this is a dp table - "dynamic programming" table
limpid smelt
oh you mean this is pseudocode for the strings code you were doing with filter?
fiery cosmos
that pseudocode is the core of my algo
limpid smelt
I've been thinking you were trying to show me pseudocode for how to find time complexity lmao
limpid smelt
I was like we were taught to use math to find complexity xD
fiery cosmos
there are other famous DP problems/algos, eg rod-cutting probem
limpid smelt
its the pseudocode for LCS, the longest common subsequence algo
aah gotcha!
idk about the algo, we were always just made to use math to figure it out and make our own algos so we werent taught like this in college haha
but thats cool!
limpid smelt
idk about the algo, we were always just made to use math to figure it out and ma...
like permutation and combination, stats etc
true, but like we weren't told that this is the LCS algorithm or this is the blah blah algorithm etc
fiery cosmos
well, if you took an algos course they likely would have
limpid smelt
as I said different countries have different flavours of teaching ig xD
fiery cosmos
did you take algos?
limpid smelt
there is no course called algos per se
but like DSA course had various algos for like searching splitting etc
but we had to figure out the pseudocodes ourselves
fiery cosmos
yeah thats just a combined data structures and algos course
limpid smelt
we were just taught theory
fiery cosmos
largely the same
limpid smelt
yeah
fiery cosmos
what text did you use?
fiery cosmos
its ok i was just curious
limpid smelt
I think it was some book by a professor ig within the college
fiery cosmos
makes sense
limpid smelt
or maybe some text copied from IITs or BITS (tier 1 colleges)
fiery cosmos
Indian Institute of Tech?
limpid smelt
we usually never even used the books. the professor would just explain on the board, we would copy that and then go back to the room and relearn from youtube
Yeah
basically rote-learning
fiery cosmos
thats a good way to do it
limpid smelt
lol no way xD
I mean atleast for me
I cant memorise stuff to save my life. I need to understand it completely and after that its fine
but people here do have a talent for memorising stuff
fiery cosmos
i'm good at both concepts and memorization. but i shied away from math early on in life and i've been paying for it ever since
just now getting my shit together
limpid smelt
we had people memorising questions and answers from 800+ page guide books..
fiery cosmos
there are strategies, for example memory palace and others
limpid smelt
aah yeah math was my favourite subject apart from cs
fiery cosmos
its not difficult to memorize. putting all the memorized concepts into theory and mechanism is where the real comprehension comes from
theres nothing important about memorization anymore in the age of google
limpid smelt
nah I do better by understanding concepts. even though it takes more time, it will benefit you in the long run
like I still remember stuff from math in school but I dont remember a single thing from chem during college
chem is the worst for me.. its just so much of stuffing random tidbits of knowledge and then regurgitating them onto your answer sheet
fiery cosmos
i love chem its one of my favorite subjects and some of the concepts are unique and fascinating eg enantiomerism
limpid smelt
so I would get like 90+ in English, Math, Physics, CS, German and then like 60 in chem xD
enantiomerism?
limpid smelt
oh, I only remember isomers faintly, no idea what they were
are you doing UG or PG?
fiery cosmos
basically, because we exist in 3 dimensions, two molecules can have the exact same molecular formula, same weight, same everything, but nevertheless be biologically different
and the only way to tell them apart is using plane polarized light
limpid smelt
wait but then wouldnt there be multiple versions of lets say the CO2 or H2O molecules?
fiery cosmos
no, in order to be a stereoisomer it must have a chiral center, eg a chiral carbon, which requires 4 different types of atoms
well, 3 not including the chiral carbon
it forms a pyramid shape with the chiral carbon at the center
limpid smelt
aah okay so a specific shape, gotcha
fiery cosmos
its not the shape that is important, but that there are 3 different atoms attached to the carbon. if two of the 3 are the same, you cannot have chirality
haughty mountain
I hated basically any subject that was just memorization
thankfully math, physics and cs is more about concepts than "just remember these random pieces of fact"
of course it's some memorization, but a lot of stuff you can reason yourself to from other facts
fiery cosmos
chemistry is much more concepts than memorization as well
bio used to be all about memorization, but now with google the best biologists are the best with concepts, biochem, immuno, etc, eg, the people who simply memorize, their days are limited
its more now about uncovering mechanisms and understanding signalling cascades, and whathaveyou
fiery cosmos
@haughty mountain are you curious about this problem my dad had to solve in 4tran a million years ago? kind of a neat problem
haughty mountain
fiery cosmos
so basically you are given 3 points in an x,y,z coord system and have to find the area of the resultant triangle
agile sundial
isn't there just a formula for that
fiery cosmos
probably. i'm curious about the algo and how it is solved
agile sundial
pretty sure it's just "plug it into the formula"
fiery cosmos
i'll google
haughty mountain
there is a simple cross product formula for this
haughty mountain
you don't want heron here
agile sundial
there's a simpler one where you don't need distances first
lean walrus
Ah, you can do cosine theorem
haughty mountain
iirc correctly this works
|(A - B) x (C - B)|/2
lean walrus
Idk how it is actually called
haughty mountain
where x is the 2d cross product
agile sundial
it's like, abs(x0(y1-y2) + x1(y2-y1) + x2(y0-y1))/2
lean walrus
iirc correctly this works
`|(A - B) x (C - B)|/2`
You also can do it without abs() to get vector-area instead of scalar-area
agile sundial
what is vector area
agile sundial
but why 😔
lean walrus
what is vector area
Vector. Length = scalar area. Direction - perpendicular to your area.
agile sundial
ah
lean walrus
It is useful in physics, for example
If you have pressure and you want to calculate pressure force, you can do: F = S*p, where S and F are vectors
haughty mountain
there is also fun generalizations of cross product that I never bothered to learn
wedge product and something else I think
fiery cosmos
iirc correctly this works
`|(A - B) x (C - B)|/2`
so this plus calculate the 2d cross product and plug in for x?
haughty mountain
wedge product and something else I think
haughty mountain
so this plus calculate the 2d cross product and plug in for x?
what I wrote is a full formula, just written in vector form, in practice A = (x_A, y_A), B = (x_B, y_B), C = (x_C, y_C)
lean walrus
It is useful in physics, for example
It is also useful when you are integrating over some surface.
haughty mountain
and
(x1, y1) x (x2, y2) = x1*y2 - x2*y1
(I think that's the correct sign at least)
fiery cosmos
i thought each point will have 3 coords?
haughty mountain
ok, are we in 2d or 3d?
haughty mountain
the formula still works, you just need to do the full 3d cross product
agile sundial
ig i don't remember as much linalg as i should 😔
fiery cosmos
i need to learn linear
haughty mountain
does anyone have 3 points with a known area?
fiery cosmos
wolfram alpha does
fiery cosmos
i need to learn linear
and think of matrices as numbers, apparently
haughty mountain
!e second time's the charm?
def cross(a, b):
return (
a[1]*b[2] - a[2]*b[1],
a[2]*b[0] - a[0]*b[2],
a[0]*b[1] - a[1]*b[0],
)
def sub(a, b):
return (
a[0] - b[0],
a[1] - b[1],
a[2] - b[2],
)
def add(a, b):
return (
a[0] + b[0],
a[1] + b[1],
a[2] + b[2],
)
def magnitude(a):
return sum(x**2 for x in a)**.5
A = (1, 2, 3)
B = (4, 1, 5)
C = (2, 5, 3)
area = magnitude(cross(sub(A, B), sub(C, B)))/2
print(f'{area = }, {area**2 = }')
@haughty mountain :white_check_mark: Your 3.11 eval job has completed with return code 0.
area = 3.0413812651491097, area**2 = 9.249999999999998
err...
not quite
it doesn't let me re-run :<
!e
def cross(a, b):
return (
a[1]*b[2] - a[2]*b[1],
a[2]*b[0] - a[0]*b[2],
a[0]*b[1] - a[1]*b[0],
)
def sub(a, b):
return (
a[0] - b[0],
a[1] - b[1],
a[2] - b[2],
)
def add(a, b):
return (
a[0] + b[0],
a[1] + b[1],
a[2] + b[2],
)
def magnitude(a):
return sum(x**2 for x in a)**.5
A = (1, 2, 3)
B = (4, 1, 5)
C = (2, 5, 3)
area = magnitude(cross(sub(A, B), sub(C, B)))/2
print(f'{area = }, {area**2 = }')
@haughty mountain :white_check_mark: Your 3.11 eval job has completed with return code 0.
area = 5.916079783099616, area**2 = 35.0
agile sundial
it's cause the syntax highlighting
fiery cosmos
wow i guess its a lot simpler than a bunch of undergrads found it to be like 40 yrs ago
haughty mountain
in any case, the cross product formulation is quite typical
fiery cosmos
i wonder how long their code was
haughty mountain
cross products are signed areas of parallelograms
lean walrus
cross products are signed areas of parallelograms
Do they work in higher dimensions?
fiery cosmos
ugh i just found a bug in my code, f***
haughty mountain
cross products sadly don't generalize well
(exterior algebra does)
without the || you get a signed area depending how u and v are pointing relative to each other
fiery cosmos
signed area 😵💫
nope.mp4
that video about newtons fractals was really cool
and then i understood what those images you shared were
they were the limit approaching the boundary between each of the zones where the points of a grid collapse to one point or another using newtons method
of approximating when polynomials = 0
opal oriole
If want all this stuff to make sense and no seem to come out of nowhere I recommend Geometric Algebra.
Then you also don't need all the other stuff like quaternions, etc.
Simplifies all formulas when used.
fiery cosmos
rn i need to find out why my code is erroring 😦
i just don't want to post the entire thing publically
haughty mountain
If want all this stuff to make sense and no seem to come out of nowhere I recomm...
which iirc is very related to exterior algebra which I mentioned earlier
opal oriole
which iirc is very related to exterior algebra which I mentioned earlier
Kind of, but more general, by combining inner and exterior.
fiery cosmos
i'm gonna learn stats, R, linear alg. in due time
not straight away
got other fish to fry
opal oriole
The "geometric product": ab = a.b + a^b
Where . is inner, ^ is wedge/exterior.
From this lots of things come naturally, like complex numbers.
Quaternions, spinors, cross product, etc (all of those are hacks added as needed, this cleans it all up into one thing).
fiery cosmos
trivia: who coined the term imaginary number
haughty mountain
guess euler or gauss in math trivia and you're right 90% of the time
haughty mountain
Descartes, huh?
fiery cosmos
👍
opal oriole
(He was being mean)
haughty mountain
"coined as a derogatory term"
fiery cosmos
oh lol i didn't realize that
opal oriole
Gauss had the right name, "lateral numbers".
haughty mountain
complex numbers are the real real numbers
so much stuff that's ugly in real math is very beautiful in complex math
opal oriole
In regular algebra*
I always find it really cool how the algebra was basically saying to the writer that it needed another channel for the information, one was not enough.
haughty mountain
closed path integrals are zero unless they surround singularities
and if they surround singularites we can evaluate the value just by looking at the values related to the singularities
opal oriole
In geometric algebra, complex numbers make even more sense.
haughty mountain
the residue theorem is so nice
opal oriole
And are just a part of it, just one case.
haughty mountain
in unrelated news, I'm debugging checksum code written in this god awful language
haughty mountain
exapunks is a nice game, very constrained programming
fiery cosmos
wth is exapunks
agile sundial
fenix have you played nandgame
agile sundial
haughty mountain
when you are done can i share my code
you can just share it, no?
haughty mountain
👀 https://www.nandgame.com/
oh god
agile sundial
it's fun lol
haughty mountain
more than normal?
I think it was more confusion about why you didn't seem to care about having posted your name in some screenshots but then being concerned about randomly generated data
this was random data you generated, no?
which you could easily regenerate
haughty mountain
so you're indexing a list with a tuple somewhere
is this the line that causes the error?
fiery cosmos
this is my input into a function which takes two input sequences
mylist is a list of lists
fiery cosmos
yes, but i thought tuples only existed in my invalid list. i must be erroneously appending an invalid tuple into my valid list
mylist should only be valid sequences, and none should be tuples
let me print it
haughty mountain
right, track the tuple backwards through the code
fiery cosmos
hmm there's something weird going at the end of my append statements in getdata
no, that's just i and j printing
huh
brittle moat
hey, I have a leetcode question that's asking to rotate a matrix by 90*. I can hardcode it, or I can do it dynamically - but would harcoding it be a bad solution?
like I'm thinking of just adding each index by the respective amount for the rotation (so if index 2 has to move to 8, index index 2 + 6 % len(matrix)) instead of running a double loop. would companies care about that?
fiery cosmos
hey, I have a leetcode question that's asking to rotate a matrix by 90*. I can h...
imo yes
brittle moat
yeah I figured they would
fiery cosmos
im still stuck if anyone doesn't mind me DMing them
haughty mountain
im still stuck if anyone doesn't mind me DMing them
so either i or j was a tuple?
if so where did the one that was a tuple come from
if in a list or similar, when and how was a tuple added?
(printing stuff will help a lot with figuring out that)
fiery cosmos
it's saying that the list index itself is a tuple, which i don't get. i printed each individually and it was fine
also, the indices are just integers running in simple loops
i do have tuples in a different list that isn't being indexed here
and that's intentional
haughty mountain
is either i or j a tuple or no?
haughty mountain
could you post the stack trace here?
you're probably using i for something else
fiery cosmos
oh yeah, there is another loop within this loop that uses i
let me test
fixed thank you
hmm
i guess i should be able to detect empty input strings
haughty mountain
empty input should be valid
fiery cosmos
big oof
haughty mountain
it shouldn't really be a special case if the code is sensible
fiery cosmos
its SO COOL it'll run and the output file echoes the input, gives the LCS of each pair, and then returns which input strings were removed and why
fricken sweet
one thing i'm not doing is returning both/all errors if there happen to be multiple invalid chars in a single string
could easily be implemented, however
cinder cedar
Nvm I understood it
fiery cosmos
can someone help me w my algo? it's erroring when there is whitespace beyond the input strings
also i want to implement error handling to detect duplicate string names
haughty mountain
can someone help me w my algo? it's erroring when there is whitespace beyond the...
elaborate?
fiery cosmos
nevermind, i'm just going to make sure the input files are formatted properly. i implemented duplicate name detection
haughty mountain
I'm just curious what extra whitespace is causing problema
end of line?
extra stuff after the last proper line?
fiery cosmos
extra stuff after the last proper line?
this
causes a list index out of range error
haughty mountain
I would just read the whole file and strip it
f.read().strip().splitlines()
and then you have sensible lines
ignoring the case of empty lines in the middle of the file
polar current
help me pls
How do I add the creation date of this post to the footer in discord.py?
haughty mountain
for line in f.read().strip().splitlines():
y = [s.strip() for s in line.split('=')]
?
fiery cosmos
😂
haughty mountain
bronze sail
algorithmic problem
haughty mountain
fiery cosmos
what about my file = f.readlines() line
haughty mountain
what about my `file = f.readlines() line`
no need
my code would replace this
#
file = f.readlines()
for line in file:
y = line.strip().split('=')
y[0] = y[0].strip()
y[1] = y[1].strip()
the other weird thing i'm doing is, instead of detecting and removing strings with duplicate names in the input, i instead detect them later and just close the program. this is in contrast with detecting error strings where i remove them and allow the program to execute. so i'm going to now try to implement handling and removing duplicate string names and allowing the valid ones with unique names to run
haughty mountain
stripping the leading/trailing whitespace of the file removes a bunch of issues
fiery cosmos
actually this brings up another potential error: where the two strings are exactly the same to begin with in the input
ehh im running up against having not made this modular enough to begin with
currently the only error strings printed to out are because they had disallowed chars
so i can't handle strings with duplicate names in the same way 🤦♂️
oh. i can just make different folders for why each are invalid
haughty mountain
modularity is helpful yes
e.g. you can easily to things one step at a time
- read (name, string) pairs
- filter out ones with disallowed characters
- handle duplicates
fiery cosmos
oh god. what if there are duplicate names and each one has disallowed chars
i guess it doesn't matter, it'll get filtered out one way or another
haughty mountain
decide what behavior you want
imo just raising an exception and exiting the program on bad input is sensible
rather than trying to salvage something from the broken input
i.e. detect error and bail the program with a message
fiery cosmos
haughty mountain
do the easy thing, if users give you bad input you scream at them
fiery cosmos
lmao
yeah i was trying to implement a custom exception but it got too complicated handling every different error possibility
and i am not going down that road now as i do not have the time
haughty mountain
do the easy thing, if users give you bad input you scream at them
it's better than gracefully failing imo, have the user actually fix their stuff
rather than accept bad input
fiery cosmos
you're probably right
tight wind
May I get some help with this problem please?
glossy breach
this is quite literally minimum spanning tree
with 0-1 weights you could use 2 dfs to solve it in O(V+E)
sonic basin
anyone online?
Anyone got any suggestions regarding what a beginner python programmer should be doing? I started self-learning python three days ago.
steel matrix
hello am complete beginner i found an article with a program that i was interested in for automatic microsoft rewards points with python. Pero I don’t know what to do for running it or whatever. would appreciate guidance .
fiery cosmos
hello am complete beginner i found an article with a program that i was interest...
Sorry, we can't help with things that would be against sites' terms of service 
fiery cosmos
Anyone got any suggestions regarding what a beginner python programmer should be...
book crash course in python or EdX python courses by David Joyner
lunar jacinth
guys
halcyon plankBOT
Hey @lunar jacinth!
agile sundial
prim's
fiery cosmos
can someone remind me how to do a batch file
open a text doc, save all the python commands, save as something.bat, run from command line?
i remember this being annoying because it wants to exist as a text file on windows
ah i got it. open notepad > save as type "all files" > somename.bat and you're good
fiery cosmos
open a text doc, save all the python commands, save as something.bat, run from c...
that should do it. e.g. test.bat contents: python -c "print('hello')", run it like .\test.bat or .\test
yeah i got it. just had to remember to open notepad and save as type 'all files' and name it .bat. i don't think right clicking somewhere and hitting create new text file would work, as it already has type .txt
you can right click and create a .txt and just change the extension to .bat
does that work on windows? feel like i've tried that. as an aside: i want to run this bat but have all the outputs go elsewhere (not my current directory), how do?
yeah, and you can use cd on a line in the bat file to change the current working directory, like bat cd ../some/other/folder python -c "import os;print(os.getcwd())"
or you can do that stuff in python https://note.nkmk.me/en/python-os-getcwd-chdir/
fiery cosmos
yeah, and you can use `cd` on a line in the bat file to change the current work...
and this'll work if the input files are in the folder where i'm running powershell?
eg the powershell cwd
when you run the bat file from powershell the cwd is wherever you are in powershell, so yes? The python script will see that as the cwd already
yeah i'm just wondering how changing the cwd within the .bat doesn't throw python's ability to see the input files
fiery cosmos
agile sundial
is there only one algorithm with 4 russian authors
hybrid epoch
https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm
Oh, that's pretty cool 😎
fiery cosmos
Hi I am searching for help
jolly mortar
@crystal tundra hey, please keep all aoc related questions in #1047673173447020564
crystal tundra
oh my bad i didn't see
fiery cosmos
guys i have a really basic algorithm question:
Suppose you have n items and you iterate through them all - this would be known as an O(n) time complexity.
If you used nested for loops iterating through n - this would be known as O(n^2) time complexity.
But, what if you have two separate for loops (not nested) what would the time complexity be.
Also, if we only iterated through half of n what would the time complexity be?
agile sundial
n, n
languid heron
https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm
they werent russians.
agile sundial
answers for your two questions. O(n) and O(n)
fiery cosmos
oh, but what if you only iterate through half of n how would it be O(n)?
agile sundial
because the definition of O(n). n/2 is only a constant factor different from n
fiery cosmos
because the definition of O(n). n/2 is only a constant factor different from n
thanks, so basically O(n/2) is equal to O(n)
its better to just say O(n)
agile sundial
yes. most people would say O(n/2) is incorrect
keen hearth
oh, but what if you only iterate through half of n how would it be O(n)?
An intuitive way to think about O(n) runtime is that if you double the size of the input the runtime doubles.
Whether the exact runtime in seconds is given by the formula n or 2*n (or k*n for any k > 0), if you double the size of n then you double the runtime.
O(n^2) means if you double the size of the input you quadruple the runtime.
O(n^3) means double input size -> multiply runtime by eight.
And so on.
O(2^n) means if you add 1 to n, then you double the runtime.
keen hearth
thanks, so basically O(n/2) is equal to O(n)
Yeah, formally O(f) represents a set of mathematical functions, and the set O(n) is literally equal to the set O(n/2) (they contain the same functions).
exotic nexus
Hey guys, im trying to create a face detection algorithm
import cv2
# step1: read the image
image = cv2.imread(r"C:\Users\awais\OneDrive\Documents\pythonProject\test1.jpg")
# step2: converts to gray image
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
# step3: get HOG face detector and faces
hogFaceDetector = dlib.get_frontal_face_detector()
faces = hogFaceDetector(gray, 1)
# step4: loop through each face and draw a rect around it
for (i, rect) in enumerate(faces):
x = rect.left()
y = rect.top()
w = rect.right() - x
h = rect.bottom() - y
# draw a rectangle
cv2.rectangle(image, (x, y), (x + w, y + h), (0, 255, 0), 2)
# step5: display the resulted image
cv2.imshow("Image", image)```when i run this code the image doesnt load and a window opens and closes for half a second
any reason why?
gentle loom
dict = {
k: {i: {j: {} for j in range(26)} for i in range(26)} for k in range (26)
}
word = "bat"
def index(word):
indices = []
for i in word:
letter_index = ord(i) - ord("a")
indices.append(letter_index)
return indices
def searchDict(dict, word):
indices = index(word)
found = False
if dict[0][indices[0]] == word[0]:
if dict[0][1][indices[1]] == word[1]:
if dict[0][1][2][indices[2]] == word[2]:
found = True
return found
def insertDict(dict, word):
indices = index(word)
print(indices)
dict[0][indices[0]] = word[0]
dict[0][1][indices[1]] = word[1]
dict[0][1][2][indices[2]] = word[2]
return
def deleteDict(dict, word):
indices = index(word)
dict[0][indices[0]] = -1
dict[0][1][indices[1]] = -1
dict[0][1][2][indices[2]] = -1
return
insertDict(dict, word)
hey guys, im trying to implement a 3 layered dictionary in python, and im trying to use my insert function. my friends say it looks right but i get this error:
TypeError: 'str' object does not support item assignment
for the line
dict[0][1][indices[1]] = word[1]
i should also mention that this is a program for 3 letter words only
i am so stumped, if someoe could help i would appreciate it
lusty swallow
regex no work me sad
haughty mountain
regex no work me sad
what would +{6,13}+ even mean?
the first + is a repeat for the character class before
the rest is just nonsense as it's written now
lusty swallow
what would `+{6,13}+` even mean?
- means searching for the same criteria multiple times right ?
so I'm applying + for the previous term and the next one
haughty mountain
the same thing before, one or more times yes, but it applies to things like characters or groups
haughty mountain
so I'm applying + for the previous term and the next one
{6,13} is 6 to 13 repetitions of something
so you have 3 "repeat this thing" in a row
lusty swallow
I was practicing regex for this question
lusty swallow
`{6,13}` is 6 to 13 repetitions of something
so it's not checking for the length?oo
haughty mountain
I was practicing regex for this question
fwiw, doing this with a single regex feels like a terrible idea
I wouldn't try to encode that as a regex
lusty swallow
fwiw, doing this with a single regex feels like a terrible idea
do I split it into multiple regex statements ?
haughty mountain
I wouldn't even do this with regex tbh
of course you could do it with a few regexes though
but feels like a roundabout way of doing things
versed charm
i want to iterate over a random integer sequence of lenght N based on a seed, and do the same iteration in reverse at the same time... is that possible without consuming O(N) memory?
the way i am currently doing this: ... is there a way to do this with generators smh?
`from random import randint, seed
seed("No Generator")
N = 10
g = [randint(1, 21) for i in range(N)]
for i in range(N):
at_forward_number = g[i]
at_reversed_number = g[-i-1]
print(at_forward_number, at_reversed_number)`
haughty mountain
technically, but not without digging into the actual random number generation
at least that's my best guess
of course you could be silly and re-generate the sequence a lot to accomplish O(1) memory, but the time complexity would be terrible if you did that
versed charm
technically, but not without digging into the actual random number generation
yeah was thinking as well, because you can "fast forward" and reverse determinsitic random number generators?
agile sundial
yes, by saving the seed at one point in time. then generate a bunch of numbers, then set the seed back to the start
haughty mountain
I'm doubtful rngs are reversible in general
versed charm
yes, by saving the seed at one point in time. then generate a bunch of numbers, ...
omg yes, ..., i think i found a solution now
haughty mountain
that's the very terrible time complexity one I mentioned
basically O(n) to get the element at index n
agile sundial
yeah. is memory really that big of a problem
versed charm
yeah .. I am doing "premature optimization" again. if the seed is not that important, this here works:
SEED = "SOLUTION"
N = 10
for i in range(N):
seed(SEED+str(i))
at_forward_number = randint(1, 21)
seed(SEED+str(N-1-i))
at_reversed_number = randint(1, 21)
print(at_forward_number, at_reversed_number)```wait, why would that work?
the seed isn't a way to jump to some point in the sequence
versed charm
wait, why would that work?
the output has the desired property for me, one goes forward one goes backward:
5 9
21 21
11 14
1 13
1 4
4 1
13 1
14 11
21 21
9 5
haughty mountain
oh wait, I'm dumb
you're only generating one thing per seed
at this point you could just do hashes or something and it would probably work comparably well
fiery cosmos
How do I learn to understand algorithms and read them from code when prompted in exams or quizzes?
Usually I defer to running the code and testing various things in a interpreter, but exams ask for a breakdown on paper with no access to a interpreter. Sometimes they can be very complex (modifying sets of sets etc).
How could I learn to think algorithmically?
versed charm
isnt that basically how minecraft can generate chunks independently from another? with a base seed? so you can teleport to different spots on the map, without having to generate the whole map?
haughty mountain
idk how minecraft rng works
fiery cosmos
isnt that basically how minecraft can generate chunks independently from another...
The world is generated with a base seed yes, the same base seed generates the same chunks always
Even if you teleport halfway accross the map, anyone accross time would see the same chunks you find there
haughty mountain
but it needs to be more complicated than that if you want things to line up at chunk borders, no?
versed charm
but it needs to be more complicated than that if you want things to line up at c...
this is where random noise (smooth) hacks come in ... If I would need to code it: every chunk has its own seed based on its coordinates mixed smh with the base seed, and then also generates the borders to the surrounding chunks, to smooth things out smh
haughty mountain
that doesn't really explain things either, what if I start at two faraway points and generate chunks inwards, then when they meet things there will be some significant clashing going on
or maybe they generate things far enough away for this not to matter and the smoothing actually to kick in
feels like this is at risk of making worlds dependent on exploration order though
I'm assuming that's not the case in reality?
versed charm
I'm assuming that's not the case in reality?
yeah its cool to think about solutions to that "jumping" around a randomly generated infinite map, and it does not depend on exploration order, as you said. generating points and interpolating between them is the solution imo
toxic dome
im trying to do a selector in python, i mean i have a school project using a raspberry
haughty mountain
#microcontrollers is probably a better fit for this question @toxic dome
oak lintel
can someone explain linear probing??
opal oriole
feels like this is at risk of making worlds dependent on exploration order thoug...
It does not depend on exploration order and uses a smooth deterministic noise function, which just takes some x, y and spits out a height (as the most simple version, Minecraft does a lot more than that now).
There are several seed reverse engineering tricks. Most use specific structure locations to simplify the problem.
Same idea also used for infinite tiling of textures in graphics in general. Pretty much all comes from Perlin Noise originally made for Tron to texture their models (giving the unique "Tron look" where everything has these infinite random patterns that are clearly generated looking).
cunning mirage
https://pastecord.com/jajuheryho
Can anyone solve this amazon DSA question?
fiery cosmos
hi, can anyone help me with any counterexample that would prove that the algorithm is wrong?
fiery cosmos
Dijkstra's algorithm fails in graphs with negative edge weights.
the algorithm never ends if the graph contains at least one negative cycle. By a negative cycle, I mean a cycle that has a negative total weight for its edges.
The Floyd-Warshall algorithm is used to detect the presence of negative cycles.
minimum = -10 , new_edge_weights = a-b= -5, a-c = -8, b->c -20
I think dijkstra's algorithm wont work with this example.
halcyon plankBOT
:incoming_envelope: :ok_hand: applied mute to @fiery cosmos until <t:1669981256:f> (10 minutes) (reason: duplicates rule: sent 4 duplicated messages in 10s).
The <@&831776746206265384> have been alerted for review.
somber raft
hello
How is this expected to be false?
abbzzca
babzzcz
switch first A with second B
babzzca
replace last a with the first z
babzzcz
same concept as this image except with 2 operations instead of 3
this is a leetcode problem btw
hybrid epoch
For LC 1657, operation 2 states: transform every occurrence of one existing character into ANOTHER existing character, and do the same with the other character.
Meaning that you can't apply operation 2 on a single letter, e.g.:
babzzca -> babzzcz
Instead you must choose at least two letters to do the transformation, e.g.:
babzzca -> babzzac
vocal gorge
hi, can anyone help me with any counterexample that would prove that the algorit...
if the graph has negative cycles, there just isn't a minimum length of a path (you can circle around the cycle forever, decreasing the path length every cycle)
but also, I think it just doesn't produce the same results in general. Adding x to each edge means a path's length gets increased by its number of edges. This can lead to different paths becoming best, e.g.
a-b-(2)-c
| |
d - e - f
# Shortest path from a to c is a-b-c, of length 3. Subtract 1 from each length, and then the path a-b-c has length 1 (because the b-c edge is length-1 now), whereas the other path has length 0.
paper tapir
Hello
Here my problem
x = 1
y = 3
z = 1
while (x != 1 or x!= 3) and (y != 1 or y != 3) and (z != 1 or z != 3):
print("ok!")
This code loops as it is not supposed to. How to exit the loop with those values ? However, this code doesn't work because it create a infinite loop...
vocal gorge
(x != 1 or x!= 3)
this is always True for any x. same for the other two.
lusty swallow
REGEX bulli me:
import re
inp=input()x`
tup=tuple(map(str,inp.split(',')))
i=0
while i<len(tup):
x=tup[i]
a=re.search("[0-9a-zA-Z\$#@",x)
if a!="None":
break
else:
i=i+1
if a!="None" and len(tup[i])>=6 and len(tup[i])<=12:
print(tup[i])
else:
print("invalid")
INPUT:
ABd1234@1,a F1#,2w3E*,2We3345
ERROR:
fiery cosmos
that regex character group has no matching ]
lusty swallow
oo ty
raven crag
how do i find the shortest path between two nodes in an undirected tree
what is the fastest solution in python
does anyone have any idea
foggy lynx
Hi there--can someone help me with a cvxpy objective function formulation?
I'm trying to do the following: compute the tax impact of a portfolio transaction. That is, my x variable is my final portfolio allocation. If x is lower than my current holding (a vector), I want to incur a tax penalty or benefit (another vector). But apparently, cp.max(data['pct_portfolio'] - x, 0) * data['tax_benefit'] doesn't work
How exactly should I formulate this?
raven crag
i need some help with this proble,m
agile sundial
how do i find the shortest path between two nodes in an undirected tree
weighted?
haughty mountain
how do i find the shortest path between two nodes in an undirected tree
there is only one path
if there are more it's not a tree
agile sundial
oh tree, yeah
exotic nexus
@hexed gust i missed your answer regarding a question previously, can i ask here?
fiery cosmos
also will increasing the edge weights have an effect on Dijkstra's algorithm when finding the optimal route?
hexed gust
<@287800670441046018> i missed your answer regarding a question previously, can ...
Sure if you still have the question. I don't really recall what it was.
limpid smelt
ax.plot(x, y, **{'linestyle': 'None','color': 'lightsteelblue', 'marker': 'o'}) for just dots
not sure about the line through the middle is that the regression line?
you can try this
visual zephyr
limpid smelt
okay, nice!
hexed gust
limpid smelt
exotic nexus
Sure if you still have the question. I don't really recall what it was.
Alright gimme a moment
exotic nexus
Sure if you still have the question. I don't really recall what it was.
import PIL.ImageDraw
import face_recognition
# load the image and convert it into numpy array
image =face_recognition.load_image_file("images/sample_image.jpg")
#find all the faces in the image
face_locations = face_recognition.face_locations(image)
no_of_faces = len(face_locations)
print(no_of_faces)
#draw the rectangular image on the face
pil_image = PIL.Image.fromarray(image)
for face_location in face_locations:
top,right,bottom,left =face_location
draw_shape = PIL.ImageDraw.Draw(pil_image)
draw_shape.rectangle([left, top, right, bottom],outline="red")
#displat the image
pil_image.save("images/output_image.jpg")
pil_image.show() ```yeah i was asking if this was a hog algorithm or not and you said the code is imported from 'face_recognition'
but when i run it on my pc the code works
urban kiln
https://atcoder.jp/contests/abc280/tasks/abc280_d
what type of math should i learn to understand and be able to solve problems like this one
I tried doing something with binary search and math.factorial but i think calculating the factorial wasnt even necessary since n could be up to 10^12 and calculating 10^12! would take way too long
fiery cosmos
I think the solution for N is the highest prime factor of K, since then N! will be a multiple of that factor and every other factor below it, times some other stuff but that doesn't change it being a multiple.
So look into factor finding I think 
fiery cosmos
I tried doing something with binary search and math.factorial but i think calcul...
^ ping since it's been a bit 
jade mountain
I made like a simple 11 line python encryption software without modules challenge to decrypt it without the key password
Dm me to get the encryptor code
And the challenge text
sharp junco
Hi guys
How do i get my python code to linear time
I have my problem solution already
But it really isn’t that efficient
If anyones willing to help, would really appreciate it
haughty mountain
I think the solution for N is the highest prime factor of K, since then N! will ...
it's not quite right
e.g. let k be some power of 2
but for sure the prime factorization is the critical thing to solving this in reasonable time
you would need to check for every unique prime factor, how far do I need to go to get that many factors
but that's fairly easy
sharp junco
How do i get my python code to linear time
@haughty mountain any thoughts?
fiery cosmos
e.g. let k be some power of 2
oh, true. K = 8 would give N = 2
urban kiln
I think the solution for N is the highest prime factor of K, since then N! will ...
ah oh
ty
haughty mountain
you would need to check for every unique prime factor, how far do I need to go t...
urban kiln
as discussed a bit above, it's not quite the solution
oh i didnt see your response
wait ill read it
ah oh
ty
hexed gust
yeah i was asking if this was a hog algorithm or not and you said the code is im...
What I mean is, it's using code from the facial_recognition package. Which looks to be built from http://dlib.net/. So no, it uses deep learning (CNN) instead of HOG. HOG is more for mobile applications as it is more light weight.
fiery cosmos
Is it worth learning how to implement more complex data structures like graphs or how/when to use them?
haughty mountain
I pmed my code
why not just post it here?
sharp junco
why not just post it here?
i dont feel comfortable sharing my code with 1000+ ppl here
haughty mountain
why?
sharp junco
if you dont wanna help its fine
haughty mountain
I'm fine with helping, but I don't want to help over dm
sharp junco
interesting
ill make a help channel then
are you fine with that
or is that a problem
haughty mountain
fwiw you're missing a big insight about the problem
is it ever useful to check longer sublists?
sharp junco
so you read the problem already
haughty mountain
yes
sharp junco
is it ever useful to check longer sublists?
thought so b4
why wouldnt it
is that making my code slower
haughty mountain
why wouldnt it
say I have a sublist L with some min and max, can adding an element make the minimum larger or the maximum smaller?
sharp junco
minimum larger
sharp junco
if the sublist L's max has a given limit, adding the element would automatically have to go the min
or vice versa
but then that wouldnt make sense thinking about it again
sharp junco
Yeah
haughty mountain
if I add another element to a list I could only make the minimum smaller
similar for maximum
sharp junco
True
haughty mountain
so I want the smallest possible list
sharp junco
Yea\
haughty mountain
in this case length 2 is the smallest allowed
checking anything larger is useless
sharp junco
then 500000 is the max
yeah
so how can i optimize
this to be more efficient
time limit is 3 seconds
for any input given
i get this when doing a large input
11 seconds is way to much
right
haughty mountain
what about implementing the thing we just talked about?
sharp junco
So we should eliminate checking anything bigger then 2 here? And have the max smaller while the min being larger?
haughty mountain
just the first part
just check all sublists of length 2
not all sublists like you're doing now
sharp junco
Seems like it’s just a matter of removing a line
Or multiple for that matter
Not really sure which ones
Would removing the second if statement be sufficient?
sharp junco
is this better? @haughty mountain
haughty mountain
no
you're still processing way more stuff than you need
you aren't checking only the length 2 sublists
the thing you have there is also just wrong
it always answers 0 afaict
sharp junco
you're still processing way more stuff than you need
ohh
mind helping me with this? ive been at it for awhile but its not clicking
like i get the objective
here
but like i dont rlly see how i should execute
or go about it
sharp junco
you aren't checking only the length 2 sublists
what your saying does make sense tho
i kinda get it ig
haughty mountain
start by iterating over all sublists of length 2, then the rest should be trivial
sharp junco
start by iterating over all sublists of length 2, then the rest should be trivia...
im having trouble with the first part of that
sharp junco
@snow arch here it is
if you can, or anyone else can
would really appreiciate the help
snow arch
im not sure about it but if im u i will take a look at threading or multiprocessing
haughty mountain
im not sure about it but if im u i will take a look at threading or multiprocess...
zero need for that here
there is a very simple O(n) solution
thorn sable
``que = []
front = 0
rear = 0
arr = []
ajmat = [[0 for i in range(20)] for j in range(20)]
b = []
p = 0
def display():
print("BFS of Graph: ")
for i in range(n):
print(arr[i])
def insert(val):
x = front
que[x] = val
front += 1
def Del():
rear += 1
return que[rear-1]
def unVisit(val):
for i in range(front):
if(val==que[i]):
return 0
return 1
def bfs(i):
j = 0
if(front==0):
insert(b[i])
for j in range(n):
if(ajmat[i][j]=='Y'):
if(unVisit(b[j])):
insert(b[j])
m = Del()
arr[p] = m
p += 1
return 0
if name == "main":
n = int(input("Enter the number of nodes in a graph: "))
for i in range(n):
b1 = input("Enter vlaue for node "+str(i+1)+" :")
b.append(b1)
print("Enter the value in adjancency matrix in from of 'Y' or 'N'\n")
print("If there exits an edge between two vertices than 'Y' otherwise 'N'\n")
for i in range(n):
print("\n")
print(b[i])
for j in range(n):
v = input()
ajmat[i][j] = v
print("\n\n")
for i in range(n):
bfs(i)
display()''
I have declared front = 0 gloally but why is it showing UnboundLocalError: local variable 'front' referenced before assignment
haughty mountain
because of how globals work in python
in insert you reassign front, which means python will ignore the global front in that function
you could use the global keyword in this case, but having mutable globals isn't great design
stuck nebula
Write a function named get_first_name that asks the user to enter his or her first name, and returns it.
method 1
def get_first_name():
first_name = input('Enter your first name: ')
return(first_name)
print("Your first name is: ", get_first_name())
method 2
def get_first_name():
first_name = input('Enter your first name: ')
return(first_name)
get_first_name()
Which method would work i tried both it works and counts towards my HW assignment i think but i wrote it different way twice to get my brain from thinking more.
sharp junco
there is a very simple O(n) solution
I'd be very open to that
mind helping me out with it?
like ive been trying for hours
but im just stuck
please would really appreciate it
haughty mountain
I'm not going to write code for you
the idea is just to iterate over all sublists of length 2 and get the difference
smallest difference is the answer
sharp junco
I'm not going to write code for you
Yeah np
was just wondering like
would I have to remove any of my existing code? or
.
here it is
haughty mountain
I would scrap it and start new, write code to iterate over all adjacent pairs
a = [...]
n = len(a)
iterate over
a[0], a[1]
a[1], a[2]
a[2], a[3]
...
a[n-2], a[n-1]
stone cipher
would calling the max function on a segment tree if l was 0 and r was equal to N have the same time complexity as calling the max function on an array normally?
sharp junco
I would scrap it and start new, write code to iterate over all adjacent pairs
i scrapped it but
this isnt any better
the efficiency still sucks
please help
i feel like i was close with
my code here
the run time was just bad
haughty mountain
this isnt any better
how is it not any better?
haughty mountain
for what size input?
halcyon plankBOT
Hey @sharp junco!
haughty mountain
500000
sharp junco
yeah
haughty mountain
!e
import time
import random
n = 500000
a = [random.randint(0, 1000000000) for _ in range(n)]
start = time.time()
ls=[]
for i in range(len(a)-1):
x = a[i] - a[i+1]
ls.append(abs(x))
print(min(ls))
end = time.time()
execution_time = end - start
print(f'Elapsed: {execution_time}')
@haughty mountain :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 292
002 | Elapsed: 0.1945819854736328
They are including the input() into the time elapsed. Hence the longer time.
haughty mountain
(+ some code to generate random data to run on)
sharp junco
<@133944101929222144> :white_check_mark: Your 3.11 eval job has completed with r...
also press the 3.10 button, want to see how fast it runs in python 3.10
hexed gust
If you are just timing the for loop, it's near instant on smaller input*.
haughty mountain
If you are just timing the for loop, it's near instant on smaller input*.
I mean, it should be
but sure, you can include parsing time as well
hexed gust
No, I mean it's including the time it takes for you to type the input in lol
haughty mountain
!e
import time
import random
n = 500000
a = [random.randint(0, 1000000000) for _ in range(n)]
s = ' '.join(map(str, a))
start = time.time()
a = [int(x) for x in s.split()]
ls=[]
for i in range(len(a)-1):
x = a[i] - a[i+1]
ls.append(abs(x))
print(min(ls))
end = time.time()
execution_time = end - start
print(f'Elapsed: {execution_time}')
@haughty mountain :warning: Your 3.11 eval job timed out or ran out of memory.
[No output]
interesting
haughty mountain
lol, too much memory I'm guessing
sharp junco
also can you check if this alternative sol works as well
import time # import time module
arr = list(map(int, input().split(" ")))
# sort the list in ascending order
arr.sort()
# initialize the minimum and maximum elements we have seen so far
min_ele = arr[0]
max_ele = arr[1]
# initialize the smallest difference between the maximum and minimum element
# in any sublist of length at least 2
diff = max_ele - min_ele
# iterate through the list, starting from the second element
for i in range(1, len(arr)):
# update the minimum and maximum elements
min_ele = min(min_ele, arr[i])
max_ele = max(max_ele, arr[i])
# update the smallest difference
diff = min(diff, max_ele - min_ele)
# print the smallest difference
print(diff)```has to run within 3 sec
largest input 500000
haughty mountain
that will give the wrong answer
haughty mountain
No, I mean it's including the time it takes for you to type the input in lol
it does, but I hope they're not typing stuff by hand...
haughty mountain
it's computing totally the wrong thing
hexed gust
it does, but I hope they're not typing stuff by hand...
Yea, I assume the 3 second limit doesn't include the time it takes you to manually input lol.
I was just stating that to clarify any confusion with timeit
haughty mountain
you can just send a file to stdin
haughty mountain
it's not like a single bug, the whole thing there is just wrong
this thing computes the right result
and it should easily run fast enough
halcyon plankBOT
Hey @sharp junco!
sharp junco
alright, makes sense. also how do i use a txt.file , such as this one as input in pycharm
i want to use it to test this problem
halcyon plankBOT
Hey @sharp junco!
haughty mountain
you could read the file from python
sharp junco
yeah was wondering how id do that
haughty mountain
or send it to stdin, though idk how to do that in pycharm
sharp junco
mind if i share screen quickly
haughty mountain
yeah was wondering how id do that
string = open('some_file.txt').read()
i have the IDE open
haughty mountain
what's wrong with just sharing code or (maybe) screenshots
sharp junco
so we can step by step approach the testing of the file?
i mean sure
but idk i thought sharing my screen would be faster
if ur not comfortable with that its fine
haughty mountain
```py
string = open('some_file.txt').read()
```
you can just read a file with the data
instead of using input()
sharp junco
OH SHOOT
bro remember the original code that i had, from the very beginning
i added one line and i think its linear time now
wanna see
haughty mountain
I'm assuming it's close to equivalent to this
sharp junco
!e
```py
import time
import random
n = 500000
a = [random.randint(0, 1000000000...
if possible, try testing it here with the 500000
haughty mountain
that one line is equivalent to the code I mentioned
idk why you have the rest there still
the loops and stuff
sharp junco
what should i remove
can you use the e! command one more time with the code
just wanna see if how fast it is
with 500000
haughty mountain
do you not see that the loop part is still quadratic?
I'm saying this is a complete solution
import time # import time module
arr = list(map(int, input().split(" ")))
result = min([abs(a-b) for a,b in zip(arr[1:],arr[:-1])])
print(result)
that second line is the same logic as
sharp junco
i get an error with the sample inputs
sharp junco
I'm saying this is a complete solution
```py
import time # import time module
ar...
this doesnt work
haughty mountain
!e
import time # import time module
s = "10 6 9 1"
arr = list(map(int, s.split(" ")))
result = min([abs(a-b) for a,b in zip(arr[1:],arr[:-1])])
print(result)
@haughty mountain :white_check_mark: Your 3.11 eval job has completed with return code 0.
3
yes it does
halcyon plankBOT
@sharp junco :x: Your 3.11 eval job has completed with return code 1.
001 | Traceback (most recent call last):
002 | File "<string>", line 3, in <module>
003 | EOFError: EOF when reading a line
your loops are quadratic time
haughty mountain
where?
sharp junco
min([abs(a-b) for a,b in zip(arr[1:],arr[:-1])])
haughty mountain
I said that one line is a solution to the problem
everything after it, the loops and stuff, is still quadratic and doesn't need to be there
haughty mountain
!e
```py
import time # import time module
s = "10 6 9 1"
arr = list(map(int, s.s...
i.e., this is a full solution
sharp junco
so this is all i need
haughty mountain
as is this
both linear
that snippet you just posted and the code I replied to just above are essentially the same
sharp junco
how similar, because i got it as an example solution, and my solution cant be exactly same, otherwise the code will be flagged
haughty mountain
oh, is none of this your code? or is some of it yours?
sharp junco
the very first was mine
fiery cosmos
anyone can help
haughty mountain
this was mine
so use that?
I didn't say exactly the same
but they do basically the same thing in the end
sharp junco
Are you familiar with MOSS
fiery cosmos
if x>283 & y>200 & x+w<356 & y+h<280 : #x+w<<420 & y+h<<330
print("x değeri:",x," y değeri:",y," x+w degerı:",x+w , " y+h değeri:",y+h)
color="undefined"
if hue_value <9:
color="red"
elif hue_value<85:
color="green"
elif hue_value<120 :
color="blue"
print("deger:",pixel_center)
but but i get value when it doesn't enter the loop
x değeri: 538 y değeri: 61 x+w degerı: 575 y+h değeri: 115
deger: [ 13 25 145]
somewhat
sharp junco
```
if x>283 & y>200 & x+w<356 & y+h<280 : #x+w<<420 & y+h<<330
...
I literally said
We almost done
fiery cosmos
oh
sharp junco
somewhat
Know of any websites that do something similar
Want to run my code through it
Or itself, if possible
haughty mountain
why? just submit your own code and you should be fine
fiery cosmos
help_
haughty mountain
```
if x>283 & y>200 & x+w<356 & y+h<280 : #x+w<<420 & y+h<<330
...
this is not a algo question
fiery cosmos
what type of this question
haughty mountain
ask in #python-discussion or open a help thread and maybe someone can help
sharp junco
!e
```py
import time
import random
n = 500000
a = [random.randint(0, 1000000000...
you tested 500000 here, how do i do that using the ide
or even here too, im open to it
seems like u did !e
but idk how it works
just to see use it for this code here
haughty mountain
it just executes code
sharp junco
!e
```py
import time
import random
n = 500000
a = [random.randint(0, 1000000000...
just trying to do what u did here
with my code here
i understand it just executes the test case
sharp junco
!e
```py
import time
import random
n = 500000
a = [random.randint(0, 1000000000...
is explaining how u tested 500000 here that bad
Alright
limpid smelt
No problem!
sharp junco
No problem!
thanls!!!
haughty mountain
is explaining how u tested 500000 here that bad
I just generated random data and ran the code
what is there to explain?
you have my code for generating the test case right there
haughty mountain
!e
import time
import random
n = 500000
a = [random.randint(0, 1000000000) for _ in range(n)]
s = " ".join(map(str, a))
start = time.time()
arr = list(map(int, s.split(" ")))
result = min([abs(a-b) for a,b in zip(arr[1:],arr[:-1])])
print(result)
end = time.time()
execution_time = end - start
print(f'Elapsed: {execution_time}')
oh right, memory issues
halcyon plankBOT
@haughty mountain :warning: Your 3.10 eval job timed out or ran out of memory.
[No output]
just run the code?
like literally just put the code above in a file and run it