Is that the Handmade Hero guy? He codes incredibly fast. Quite the marvel.
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
Programming |OT| C is better than C++! No, C++ is better than C
- Thread starter Kikarian
- Start date
Is that the Handmade Hero guy? He codes incredibly fast. Quite the marvel.
I watch those videos in awe.
This is amazing.Wait... why? The last thing I'd ever want is a peanut gallery watching me write code.
Clicked a random stream, and it was a guy scrolling through a Stack Overflow thread with a perplexed look on his face, so at least it's capturing the real coding experience.
Different versions of PowerShell behave somewhat differently, and may differ a bit on a different/newer OS (such as available cmdlets).
I would expect the latter one to work, and I would have likely never tried the former.
What I would have done:
Code:
Get-ChildItem -Path C:\Users | Select-Object -ExpandProperty FullNameUltimately it wasn't as hard as I'd made it. I'd just read on their setup pages that implied Java 8 wasn't supported on Tomcat there and I needed Java 8 support due to features of it being used in my project so I had no wiggle room and went down a couple of dead ends. In the end I just used a Wildfly server and it worked no problem, kinda wondering if I just used a tomcat and deployed my WAR it would work but it doesn't really make a lot of difference to me if it's using Tomcat, JBoss, Wildfly anyway.
I stumbled upon this: http://stackoverflow.com/questions/...in-implementing-binary-search/6393352#6393352
And I have a question for people who know Pascal. Supposedly the Procedure he gives exhibits the following errors as far as implementation goes:
And... I'm not seeing the error. For 3, I don't understand how the loop can possibly exit on Low == High. There's no alternative terminating condition to (Low > High) or (Key = A[Index]). If Found is set after the loop terminates without finding the Key, then Low will always > High. Is Found somehow set in parallel to the loop exiting?
For 4 and 5, I get that the offending line is:
But I don't understand how the loop can ever reach a point where it evaluates to < 1 or > Size. REPEAT-UNTIL is a posttest loop is it not? I can see the error if it's a prettest, so Low == High == 1 will allow the loop to execute once more and access an out-of-bounds index, but I'm unsure how that case slip through a posttest.
EDIT: Okay, so. Error 3 is written incorrectly. The procedure will give a false negative, not a false positive. And Pascal has invariants on index types apparently, so Low or High ever being out of bounds will set off an error.
Code:
PROCEDURE BinarySearch (A : anArray,
Size : anArraySize,
Key : INTEGER,
VAR Found : BOOLEAN;
VAR Index : anArrayIndex);
Var Low, High : anArrayIndex;
BEGIN
LOW := 1;
High := Size;
REPEAT
Index := (Low + High) DIV 2;
If Key < A[Index]
THEN High := Index - 1
ELSE Low := Index + 1
UNTIL (Low > High) OR (Key = A[Index]);
FOUND := (Low <= High)
END;And I have a question for people who know Pascal. Supposedly the Procedure he gives exhibits the following errors as far as implementation goes:
And... I'm not seeing the error. For 3, I don't understand how the loop can possibly exit on Low == High. There's no alternative terminating condition to (Low > High) or (Key = A[Index]). If Found is set after the loop terminates without finding the Key, then Low will always > High. Is Found somehow set in parallel to the loop exiting?
For 4 and 5, I get that the offending line is:
Code:
Index := (Low + High) DIV 2;EDIT: Okay, so. Error 3 is written incorrectly. The procedure will give a false negative, not a false positive. And Pascal has invariants on index types apparently, so Low or High ever being out of bounds will set off an error.
close to the edge
Member
Gonna try to do some Rust streaming now. Bear with me, I've not done this before. I'll probably be doing a re-implementation of something I did earlier (travelling salesman problem) which broke as a result of the various changes before the language became stable.
https://www.livecoding.tv/gyrosofwar/
edit: Okay, OBS on Linux still seems to be having some issues. It kind of stops working after I've streamed for a couple of minutes.
Gonna try it on Twitch in the meanwhile..
http://www.twitch.tv/klo8/
I did very much enjoy my computer architecture courses, mostly because it demystified these "computer" things. So now I have a basic understanding of how computers work at that lowest layer. I've also done a systems programming course where we did a bunch of stuff IPC and system call stuff, that I very much enjoyed.
I think a good next step might be to learn how the kernel (and device drivers and such) works, since that's the main thing that's still kind of black magic to me. Can anyone recommend any good books on kernels?
I'm also thinking that setting up a "Linux From Scratch" install might be a great way to learn a lot about operating systems, perhaps I'll do that in the break between my summer course and the start of the semester.
Codeacious
Banned
Sadly the 4th isn't out yet (wasn't it supposed to come out last year?) and some things might not work with current(3.x+?) kernels/have to be done different. But yeah, to get a basic understanding it is still all right, even though it is missing the glory that is device trees.
Besides that there is the kernel documentation within the kernel source tree which may or may not be up to date but should touch topics beyond the drivers. And I suppose Tanenbaum's for even more general information.
Chaos
Member
There is an android developemnt course kickstarter here if anyones interested -
https://www.kickstarter.com/project...-m-developer-course-with-andro?ref=nav_search
Becase it has gone over the inital goal you will get 2 other courses for free.
https://www.kickstarter.com/project...-m-developer-course-with-andro?ref=nav_search
Becase it has gone over the inital goal you will get 2 other courses for free.
i have a question about file input/output in python and the read method.
writing a function to read one file and write to another
this gives me the following error.
NameError: global name 'fl' is not defined
but if I modify the function like so
I no longer get the error. so why do I get the name error when I try to close f1 in the first case?
writing a function to read one file and write to another
Code:
def sed (input_file, output_file):
f1 = open(input_file, 'r')
f2 = open(output_file, 'w')
text = f1.read()
print text
f2.write(text)
fl.close()
f2.close()
returnthis gives me the following error.
NameError: global name 'fl' is not defined
but if I modify the function like so
Code:
def sed (input_file, output_file):
f1 = open(input_file, 'r')
f2 = open(output_file, 'w')
text = f1.read()
print text
f2.write(text)
if text =="":
fl.close()
f2.close()
returnI no longer get the error. so why do I get the name error when I try to close f1 in the first case?
mugurumakensei
Member
You got a typo. The problem is you're calling close on 'fl' (f - lower case L) not 'f1' (f - one).
My guess is f1 is empty. fl.close never gets called since you never go into the if which means you never encounter the name issue.
well now if feel stupid. i thought it was something to do with how the read method works. Thanks.
Hey guys, check out my first site. Made in PHP, MySQL, JS and Foundation. I did this all in procedural PHP so I can have a good enough idea of what goes into making a website that I won't be confused by frameworks like Ruby on Rails (learning that next). Registration not open because I haven't set up emails yet (and haven't finished a few features).
Let me know if there are any bugs
.
Let me know if there are any bugs
.Not sure why anyone would be interested. Why kickstart something, when those kinds of courses already exist (and tons of tutorials and other resources) and are ready for purchase or even free?
Google themselves have free courses on udacity:
- https://www.udacity.com/course/android-development-for-beginners--ud837
- https://www.udacity.com/course/developing-android-apps--ud853
CornBurrito
Member
So this is perhaps more math than programming related, but I'm slowly going through Project Euler and I've just done problem 5. My code works, but it is very fucking slow. I'm not really sure what route I could take to improve the speed.
I start at 2520 because I simply reasoned that the number would have to be greater than in the (1-10's) case. Similarly I increment a by 20 because I know it has to be an even number, and that it will need to be divisible by 20. I think I could probably increase the increment but I'm not sure whats the highest I could increment it. And I also don't think I chose the best starting point either.
I do get an answer though. It just takes a bit of time. About 6-7 Mississippi seconds.
edit: Googling around I found a small hint and changed the increment to 2520. With that the thing gets the right answer in like under a second. But can someone explain this to me more mathematically because I was honestly just guessing with my starting point and increment to begin with.
Code:
"""2520 is the smallest number that can be divided by each of the numbers from 1 to 10 without any remainder.
What is the smallest positive number that is evenly divisible by all of the numbers from 1 to 20?"""
count = 0
a = 2520
while count < 21:
for i in range(2,21):
if a % i == 0:
count += 1
else:
count = 0
a += 20
print aI start at 2520 because I simply reasoned that the number would have to be greater than in the (1-10's) case. Similarly I increment a by 20 because I know it has to be an even number, and that it will need to be divisible by 20. I think I could probably increase the increment but I'm not sure whats the highest I could increment it. And I also don't think I chose the best starting point either.
I do get an answer though. It just takes a bit of time. About 6-7 Mississippi seconds.
edit: Googling around I found a small hint and changed the increment to 2520. With that the thing gets the right answer in like under a second. But can someone explain this to me more mathematically because I was honestly just guessing with my starting point and increment to begin with.
CornBurrito
Member
Yeah it is actually pretty fun to see what others have posted. Comparing code and going "holy fuck, this guy's solution is genius how did I not see that" is really great.
I checked if I posted my problem 5 solution, but I didn't. Might have it lying around on my laptop. I'll check later.
A more mathematical solution can easily be found if you work out the problem for numbers between 2 and 10 by hand.
2 is the smallest number that is evenly divisible by 1 to 2.
6 is the smallest number that is evenly divisible by 1 to 3.
12 is the smallest number that is evenly divisible by 1 to 4.
hint:
2 is the smallest number that is evenly divisible by 1 to 2.
6 is the smallest number that is evenly divisible by 1 to 3.
12 is the smallest number that is evenly divisible by 1 to 4.
hint:
Factor each number and it should be obvious
Yeah. I think I just reduced the factors by hand on paper. I tried to find my posts for other problems and there seems to be huge gaps in the answer threads. They must auto-delete some of them.
Could anyone give me a tip on how to do project Euler problem 27?
I'm on my phone so can't really post my current code sorry. I'm stumped beyond just going brute force which takes too long.
https://projecteuler.net/problem=27
I'm on my phone so can't really post my current code sorry. I'm stumped beyond just going brute force which takes too long.
https://projecteuler.net/problem=27
So, i'm at that stage where i should be sending resumes out.. and..
It's me or the requirements most companies put up, even for junior positions, are completely insane?
3y+ in every single part of their stack? How does that even happen?
It's me or the requirements most companies put up, even for junior positions, are completely insane?
3y+ in every single part of their stack? How does that even happen?
You're going to need to take a crack at it before we can help you. If you have more specific questions feel free to ask. Your teacher has laid out exactly what they want you to do, conceptually.
I'm thinking something like this (pseudocode)
Code:
class Coefficients {
int a;
int b;
}
class Quadratic {
Coefficients c;
int n;
boolean isPrime() {
/* Check for prime, sieve of eratosthenes? */
}
}
Map<Coefficients, Integer> coefficientsToNumPrimes = Maps.empty();
for (int a = -999; a <= 999; a++) {
for (int b = -999; b <= 999; b++) {
Coefficients c = new Coefficients(a, b);
int numPrimes = 0;
for (int n = 0; true; n++) {
Quadratic d = new Quadratic(c, n);
if (d.isPrime()) {
numPrimes++;
else {
coefficientsToNumPrimes(c) = numPrimes;
break;
}
}
}
}Then just find the entry in the map with the largest number of primes and multiply the coefficients.
Ok then, for the lines that say:
Is this saying to set up a for loop setting the number of loops in the parameters or what?
I've seen quite a few entry level positions ask for multiple years in areas and I was confused on that. I assume they expect people to have used those specific tools in a non-professional area. Actually ended up interviewing for a spot like that one time through some weird circumstances.
It's saying that you need to read in a number from the user. This number should be used to make your for loop execute that many times.
For example, if I had read an int "loopCount" from the user, and 10 was entered, the following loop:
Code:
for (int i = 0; i < loopCount; i++) {
/* do something */
}would execute 10 times.
CornBurrito
Member
Is there a better way to implement the Sieve of Eratosthenes in python than the way I have below?
I've read online that xrange is faster than range, but xrange doesn't have remove, and my guess is that xrange doesn't create a list while range does. I suppose it might be faster to iterate over all the numbers using xrange and then append the primes that show up to a list, instead of creating a list and then removing the non-primes?
Code:
"""By listing the first six prime numbers: 2, 3, 5, 7, 11, and 13, we can see that the 6th prime is 13.
What is the 10,001st prime number?"""
def sieve(n):
initial = range(2,(n+1))
for i in initial:
for x in initial:
if x % i == 0 and x != i:
initial.remove(x)
return initial
print sieve(110000)[10000]I've read online that xrange is faster than range, but xrange doesn't have remove, and my guess is that xrange doesn't create a list while range does. I suppose it might be faster to iterate over all the numbers using xrange and then append the primes that show up to a list, instead of creating a list and then removing the non-primes?
CornBurrito
Member
What do you mean by safe in this context?
And I do have Python 3 on my computer, but I'm mostly doing stuff with 2.7.9 right now and a lot of stuff still uses it. So I'm not in a rush to switch. I'll probably make the switch when I start doing stuff with pygame.
Chainsawkitten
Member
I'm coming from C++/Java/etc so Python may be doing these things automatically, but generally you can't delete elements in a list that you're currently iterating over.
For example, let's assume we have an array list. It has the values [0, 1, 2] in it. I want to loop through it and print all the values, and if the value is 0, remove it from the list.
So we do:
Code:
for (int i=0; i<list.length(); i++) {
print list[i];
if (list[i] == 0)
list.delete[i];
}Here's what's going to happen:
- i = 0, list[0] = 0
- We print the 0.
- We remove the 0. The list is now [1, 2].
- i = 1; list[1] = 2 (list[1] was previously 1, but it changed when we deleted list[0])
- Print 2
- We're at the end of the list. So we're done.
Actual output: 0, 2.
But it gets worse. What if we don't want to recalculate the length of the list every loop (nevermind that the list would usually store that as a variable if it's any decent implementation)? Then we do:
Code:
int length = list.length();
for (int i=0; i<length; i++) {But you didn't do a for-loop with an index like that, you used an iterator, so it shouldn't be a problem! ... except an iterator is just a more convenient way to express that (in the context of an array list, a linked list iterator doesn't use indices), so it's actually doing that under the hood.
So what do we do if we want to modify a container as we loop over it? Well, we'll have to use the iterator's special remove function, which updates it's internal index and length variables (if it's using a length variable) so that it's accurate and we can keep looping over the container without getting unexpected output or breaking things.
Now, again, I don't know Python so I don't know if it magically fixes things, but from reading your code it seems like it would be a problem, unless Python lists keep track of all their iterators and updates them whenever the container is modified (which seems like quite a lot of extra overhead).
wolfmat
Confirmed Asshole
Removing (popping) is cheaper than appending because to append, you have to put data somewhere, while popping just means you drop a reference (and return it of course, but that's irrelevant here). However, you cannot expect a list iteration result to be sane if you pop something in the middle and do not consider this later with an offset (the list has changed after all; something's gone). But the popping will always be cheaper.
Edit: The offset is the common thing to do to address concerns like "doesn't iteration break if I change the list while iterating over it?"
Know that the for-loop setup is immutable only if you're NOT using a cursor. So if you do
Code:
a=[1,2,3,4]
for i in a:
print('p',a.pop())
print(i)p,4
1
p,3
2
as the output, and are left scratching your head.
But if you do
Code:
a=[1,2,3,4]
for i in range(0,len(a)):
print('p',a.pop())
print(i)('p', 4)
0
('p', 3)
1
('p', 2)
2
('p', 1)
3
So that's something to keep in mind: Is my for-loop setup immutable or not?
Chainsawkitten
Member
Depends on how the list is implemented. Removing things in the middle of an array list is probably a lot slower that populating a new list, just appending things at the end. If it's a linked list, though, removing is O(1).
CornBurrito
Member
However my code for certain works. At the very least for finding the 10,0001st prime number. And the 6th prime number. Python allows this*. Downside is it takes a significant chunk of time. Though I've looked up other answers at this point and found a better application of the sieve
*Unless by "can't" and "unsafe" you guys more mean "shouldn't because outside this problem things might get fucked."
huh interesting. I'll keep this in mind, thanks.
wolfmat
Confirmed Asshole
Oh yes, I was of course just referring to pop(), which drops the last element. pop(0) is hardly performant here.
I guess I read it wrong because he surely has to pop somewhere inside the array.
Anyway, "initial.remove(x)" is even worse than popping because the item has to be found, which is more than rebuilding the list.
I think one should either trick one's way to performance with dictionaries (O(1) everywhere), or append to a result list.
CornBurrito
Member
Ok so I was heading in the right direction when I said I thought appending to an empty list in this case would be faster than .remove() from a huuuuuuge list.
I may also try doing this with dictionaries as you suggest, since I've barely ever used dictionaries.
wolfmat
Confirmed Asshole
Yeah we've boosted performance like crazy in a product by just NOT doing things like
Code:
if 149871 in some_array:
print('yes')
Code:
try:
if some_ref_dict[149871]:
print('yes')
except KeyError:
passThe reference dict gets value True for key 149871 before because 149871 is a number that matches some condition; in the append() scenario, you'd have appended that number to a these-numbers-match-the-condition list earlier.
This is applicable in lots of cases, and trivializes such checks.
Thanks, will give this a try.
There are some constraints on a and b.
1) b must be prime and >0, total # of b = 168
2) a > -b, and a must be odd.
PS. Now I think about it, brutal forcing it may not be that bad after all.
CornBurrito
Member
I'm feeling dumb because I can't wrap my head around this syntax.
I've marked the line I don't get in bold. I tried making changes to see what happens, and the only thing I got without breaking it completely is if I make it
is a list containing a few copies of limit-1. That number of copies being the number of primes within that limit.
I get that this creates a list. At first I thought it was something like [i for i, stuff you want i to be], such that [i for i, range(2,10)] would create a list of [ 2, 3...9]
However that doesn't appear to be the case since if I make that change I get an invalid syntax error. I also don't understand why doing something like "i for x, stuff" prints a list containing the correct number of primes, but with limit-1 taking place of the prime numbers.
Code:
def iprimes(limit):
is_prime = [False] * 2 + [True] * (limit - 1)
print is_prime
for n in range(int(limit**0.5 +1)):
if is_prime[n]:
for i in range(n*n, limit+1, n):
is_prime[i] = False
[B]return [i for i, prime in enumerate(is_prime) if prime][/B]I've marked the line I don't get in bold. I tried making changes to see what happens, and the only thing I got without breaking it completely is if I make it
Code:
return[i for x, prime in enumerate(is_prime) if prime]I get that this creates a list. At first I thought it was something like [i for i, stuff you want i to be], such that [i for i, range(2,10)] would create a list of [ 2, 3...9]
However that doesn't appear to be the case since if I make that change I get an invalid syntax error. I also don't understand why doing something like "i for x, stuff" prints a list containing the correct number of primes, but with limit-1 taking place of the prime numbers.
I'm not sure since I never used Phyton (this is Phyton, right?), but I believe it's [X for Y in Z]. X is a expression that will return the values of the list, in that case, the values of i. Y are the variables you're going to use to get these values, in that case i and prime are the indexes and the items of the is_prime array, respectively. Z is where you are getting the values from, and in that case it's enumerating the indexes and values of is_prime and putting them on i and prime. Since is_prime is a bool array, prime is a bool, and if it's true, i get's added to the list.
CornBurrito
Member
Yeah it is Python. Anyway I think you may be right. Thanks for the explanation. The strange thing is that Python does allow you to make lists via [x for y in z], so I don't know why this alternate syntax is being used.
It's a conditional generator. [x for x in some_iterable if some_condition] in this case x is only added to the list if the condition is satisfied. A generator [X for Y in Z if some_condition] nets the same result as
Code:
result = []
for Y in Z:
if some_condition:
result.append(X)that's because i is from the loop two lines above, not from the enumerator.
In addition to the other responses I think it's good to consider the statement in detail. The statement can be deconstructed into three separate parts:
Code:
i for
i, prime in enumerate(is_prime)
if primeThe first part i for creates a list of elements i. The second part i, prime in enumerate(is_prime) is a pattern match against a generator. The third part if prime is a conditional which makes sure that the values i are added to the list only if their corresponding values prime are True.
In this context, is_prime is a list of True/False values, corresponding to whether or not the numbers 0, 1, 2, ... are or aren't primes. So:
Code:
i, prime in enumerate(is_prime)
Code:
i, prime = 0, False # Binds i = 0, prime = False
i, prime = 1, False # i = 1, prime = False
i, prime = 2, True # i = 2, prime = True
...So, the syntax
Code:
[i for i, prime in enumerate(is_prime) if prime]As for the statement
Code:
[i for x, prime in enumerate(is_prime) if prime]
Code:
x, prime = 0, False # x = 0, prime = False
x, prime = 1, False # x = 1, prime = False
x, prime = 2, True # x = 2, prime = True
...
Code:
[0 for i in [0, 1, 2, 3]] # This creates the list [0, 0, 0, 0], i is not used.
[i for i in [0, 1, 2, 3] if False] # This creates the empty list [] since the conditional is always FalseSo what happens in your last case? Basically, the last value that i was set to is used. This corresponds to:
Code:
i = limit - 1 # ie. 5 if limit = 6
[i for x in [0, 1, 2, 3]] # Creates [5, 5, 5, 5]I like how these list comprehensions are implemented in Python, although they can be confusing. Especially if you want to use an if/else condition, in which case the conditional comes before the generator, ugh.
Code:
["even" if i % 2 == 0 else "odd" for i in [0, 1, 2, 3]] # Creates the list ["even", "odd", "even", "odd"]