|
Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks. |
On June 26 2017 08:14 KungKras wrote:Show nested quote +On June 26 2017 07:52 Hanh wrote:
It helps when figuring out what units are affected by aoe or hit or even visible. That's interesting. I never thought about AOE spells or visibility. Is it because you don't have to iterate through every object in the game to find what units are in an AOE when using a tree?
I don't know about the binary tree, but if it's just about figuring what's happening in the neighborhood of the unit, then you could already do that easily by just looking at the location and sweep around the unit. As a first approach that's not really bad. Most of the time, your units will be stored in some kind of grid-like discretization of the map (a 2D array basically). You can just sweep in a determined bounded small radius around your unit's cell.
If you're looking for some fancy way of accessing information on the map, you could look into quad-trees (Edit: ah I see that the quad-tree is mentioned in the previous page  ) and the like, which are designed to handle information in 2D grids.
|
Those link were amazing! Thanks :D
On June 26 2017 22:10 ZenithM wrote:Show nested quote +On June 26 2017 08:14 KungKras wrote:On June 26 2017 07:52 Hanh wrote:
It helps when figuring out what units are affected by aoe or hit or even visible. That's interesting. I never thought about AOE spells or visibility. Is it because you don't have to iterate through every object in the game to find what units are in an AOE when using a tree? I don't know about the binary tree, but if it's just about figuring what's happening in the neighborhood of the unit, then you could already do that easily by just looking at the location and sweep around the unit. As a first approach that's not really bad. Most of the time, your units will be stored in some kind of grid-like discretization of the map (a 2D array basically). You can just sweep in a determined bounded small radius around your unit's cell. If you're looking for some fancy way of accessing information on the map, you could look into quad-trees (Edit: ah I see that the quad-tree is mentioned in the previous page ) and the like, which are designed to handle information in 2D grids.
It seems like quadtrees are exactly what I'm looking for. I'm curious about how it would work in 3D space, but since I'm making a 2D one, quad trees are perfect.
|
Manit0u why rails over python?
|
|
|
On June 26 2017 23:15 ShoCkeyy wrote:
Manit0u why rails over python?
I know some Python too but there are many more job opportunities in Rails in Poland (and all over Europe, judging from the number of offers people are making me). I also like Ruby more than Python.
And no matter how reluctant I am towards Java, I believe we should switch to it (or Scala) so people can actually learn how to do enterprise stuff (and Java 8/9 isn't so bad). Doing enterprise stuff in Rails is simply unconventional.
Edit: Yay! My team lead wants to try out Scala (especially that its syntax is very close to Ruby). Keeping my hopes up.
|
We're trying to switch to Kotlin, it's a really fun language to program in :d
We're also trying to switch to using Buck, but that's not going so well... gradle...
|
anyone know why I am getting an error with this code?
it is for this problem:
https://leetcode.com/problems/remove-element/#/description
class Solution(object):
def removeElement(self, nums, val):
"""
:type nums: List[int]
:type val: int
:rtype: int
"""
nums.sort()
start=-1
end=-1
for i in xrange(0,len(nums)):
if nums[i]!=val:
if start<0:
start=i
else:
end=i
return nums[start:end+1]
I get Line 58: TypeError: range() integer end argument expected, got list.
It works fine for me on my IDE and rep.it though
FWIW not sure that solution is accepted, didn't bother checking after I got hooked onto figuring out why the heck I keep getting this error
edit:
I guessed that it might have to do with me returning a list and misreading the problem. I assume it's asking for a list but I tried returning the length of the list (which works again on rep.it and my IDE) but bugs out on leetcode with "Unknown Error".
Not sure if I inadvertently bugged out the system or if I actually have some strange error that happens to work on other environments.
|
Have you tried returning the item that was removed? That is more typical of a remove method.
Given an array and a value, remove all instances of that value in place and return the new length.
|
Well, for starters, you're supposed to do it in-place, so sorting screws that up.
But why did you not just go with the really obvious O(n) solution of walking through the list and noting all indices, then removing all elements you found with del (and returning the length of the list?)
|
On June 28 2017 06:31 Blitzkrieg0 wrote:
Have you tried returning the item that was removed? That is more typical of a remove method.
well I didn't exactly remove the item since I'm just returning a subarray so both will be returning a subarray
but no I got the same error on the site, no error everywhere else
On June 28 2017 06:43 Acrofales wrote:
Well, for starters, you're supposed to do it in-place, so sorting screws that up.
But why did you not just go with the really obvious O(n) solution of walking through the list and noting all indices, then removing all elements you found with del (and returning the length of the list?)
well I was just doing a naiive solution since I literally just barfed out whatever I thought of in 30 seconds
I did consider your type of solution for a second but thought noticing the indices would take a non constant amount of space(is this wrong?)
of course this is with my possibly naiive assumption that you are storing the indices in another list or something similar since that's what I initially thought of
I'm not super familiar with del though so maybe there is something I'm not aware of that you can do with it. I know I couldn't delete from a list while iterating through it
I sorted because I couldn't think of a way where I could do this in constant space without the list being sorted
edit:
nvm figured out how to use del to delete while iterating lol
start=0
size=len(nums)
while start<size:
if nums[start]==val:
del nums[start]
size-=1
start+=1
return nums
however I still get the same error. so I'm wondering why that is regardless of my solution
edit2:
on topic of the solution itself, if we are just looking for the length we could probably save time since I believe del is an O(n) operation?
then just use two pointers and have the element to be removed swap with the last element and decrement the end pointer. return the pointer that started from 0 since after doing all the 'swaps' the start pointer will be at the end of our 'new' list.
nvm we can do this for returning the list too with nums[0:startIndex+1]
this would be a real O(n) I believe
the solution with del is probably O(n^2)? not completely sure on the python method complexities
and I guess my original solution was O(n^2 log n) according to this https://stackoverflow.com/questions/25813774/complexity-of-python-sort-method
|
Does that actually work? If the array is [1 2 2 1] and you're removing 1 doesn't your code remove the first element and then not check the last one because it decremented the size.
|
On June 28 2017 07:24 Blitzkrieg0 wrote:
Does that actually work? If the array is [1 2 2 1] and you're removing 1 doesn't your code remove the first element and then not check the last one because it decremented the size.
The del one? Yeah it works. It checks the last one still because the list itself got shorter.
|
# @param {Integer[]} nums
# @param {Integer} val
# @return {Integer}
def remove_element(nums, val)
nums.delete(val)
nums.length
end
EZPZ
Edit: Solution for python 3
class Solution(object):
def removeElement(self, nums, val):
"""
:type nums: List[int]
:type val: int
:rtype: int
"""
return len(list(filter(lambda e: e != val, nums)))
|
Why would the first version be O(N^2 log N). You sort only once. The one where you don't sort would be O(N^2) but it doesn't work. You can use the page you linked us to test it yourself. I don't think you even tested it since it doesn't work?
Edit: Manit0u: Your python solution isn't in-place so it doesn't pass the test cases.
|
On June 28 2017 07:50 slmw wrote:
Why would the first version be O(N^2 log N). You sort only once. The one where you don't sort would be O(N^2) but it doesn't work. You can use the page you linked us to test it yourself. I don't think you even tested it since it doesn't work?
Edit: Manit0u: Your python solution isn't in-place so it doesn't pass the test cases.
I have to admit I'm not super good with Python
Edit:
I think I figured it out
class Solution(object):
def removeElement(self, nums, val):
"""
:type nums: List[int]
:type val: int
:rtype: int
"""
nums[:] = [x for x in nums if x != val]
return len(nums)
Solution accepted.
|
On June 28 2017 07:50 slmw wrote:
Why would the first version be O(N^2 log N). You sort only once. The one where you don't sort would be O(N^2) but it doesn't work. You can use the page you linked us to test it yourself. I don't think you even tested it since it doesn't work?
Edit: Manit0u: Your python solution isn't in-place so it doesn't pass the test cases.
Nvm o(n log n) since its essentially n log n (the sort) + n(walking through list)
Had a brain fart. Todays been a long day... 14 hrs of work+class :/
Also ?? It doesnt work as in it gives the error I posted about or it doesnt work as in it returns wrong values?
Not sure if you read my post above since I mentioned I tested all 3 solutions on my own IDE and rep.it where all worked fine, so yes I did test it lol.
Edit: sorry sounds sassy not intended
edit2:
ok back at home and apparently any form of slicing a list is giving me issues so doing it without any slicing worked
still dunno why
solution might be buggy but it's working as intended for me with the same test cases it's failing so idk
also the expected output is super misleading since it looks like a list for ex [2,2] when it wants the length. so returning a list is no good
@manitou I like your solution, pretty clean
it's pretty fast too, idk how mine is so much slower using 2 pointers with O(n) time and O(1) space
yours was almost 2 times faster than mine.
start=0
end=len(nums)-1
while start<=end:
if nums[start]==val:
nums[start],nums[end]=nums[end],nums[start]
end-=1
else:
start+=1
return start
@slmw
yeah that was it, thanks
the expected output confused me
|
Oh sorry, I thought you already realised where the error comes from. Return the length of the "new" array as instructed in the problem description. Not the most obvious error message...
|
On June 28 2017 08:22 dsyxelic wrote:@manitou I like your solution, pretty clean it's pretty fast too, idk how mine is so much slower using 2 pointers with O(n) time and O(1) space yours was almost 2 times faster than mine.
start=0
end=len(nums)-1
while start<=end:
if nums[start]==val:
nums[start],nums[end]=nums[end],nums[start]
end-=1
else:
start+=1
return start
Too much complexity and code branching  So many more instructions for the processor...
But seriously, pythonic list comprehensions are a big focus for optimization on their part so it's usually best to use them instead of using loops and iteration.
|
Totally unrelated, but I need to ask:
My company is desperately looking for 2 DevOps. Preferably in EU timezone so we won't call you at night (company is US based but 99% of our development happens in Poland). You'd have to do a month at our dev center here but after that you can work remote if you want to. It's mostly just server and vpn management including vmware and docker. Some CI stuff and such. Knowledge of apple servers would be a bonus (need that for one of our clients).
For more details: https://www.newshubmedia.com/careers and click on DevOps (or other things if they interest you)
If you're interested, just PM me
|
So we do job advertising now ??? We are looking for 50+ people for work in automotive also in Poland C/C++, DSP, JavaEmbedded, CAN, test automation (some of those required, not all). If anyone is intrested just let me know Unfortunetly since we work with hradware no remote working possible.
And thats what we do
+ Show Spoiler +
|
|
|
|
|
|
EPT
