its my first question here and maybe i'm just blind or extremely stupid, but i haven't found a matching answer to my problem in the past three days. Currently i'm coding an Idlegame (just to learn more about OOP and such stuff) and obviously there will be some numbers that are too big for int or long (sth like "1500000000000" (1.5E12)). So i decided to go along with floats for two reasons:
Every other type that can store way more than int or long (like BigInteger or BigDecimal) wouldn't solve my problem of shortening my shown numbers in the ui.
I like the numberformatting of float with the exponent shown at the end.
Because of some issues with the length of the outcoming number (1.50000003E12 (i think the ...03 at the end of the number comes from the Float.toString() which i use to set it onto a javafx label)) i thought about formatting them in an uniform way (in my opinion it should look like 1.23E45) so i googled a lot, read a lot, tried a lot, but still got no working solution. on SO i read multiple ways to format floats with "String.format()" and "DecimalFormat", but the outcome wasn't as expected. the float value is written completely what in fact is the opposite of that, what i wanna achieve.
So my wish would be a working solution, or a hint that would help me further to achieve my goal to only show the first three digits of a float and add the exponent.
If anything is missing that would help to find the best solution for my problem feel free to ask for it.
Related
Doing some practice coding problems in Java from Advent of Code and came across one where it asked me to find the smallest six-digit integer, combined with the leading string of iwrupvqb, that produced an MD5 hash that started with five zeros.
I found the answer to this part using the Apache DigestUtils.md5Hex function where I just brute forced through 100000-999999 and combined it with iwrupvqbuntil I got an MD5 that started with five zeros.
The answer came out to be iwrupvqb346386 creating the hash:
0000045c5e2b3911eb937d9d8c574f09
Now it's asking me to find one with six leading zeros. I've been going through pages and pages of how the md5 algorithm works, inverting MD5, etc but can't seem to figure out the problem in an equation format that would help me determine how the MD5 would be calculated based on the characters used.
I even let this loop run for like 30 mins - an hour to see if it got any hits outside of a six digit integer (because apparently there are none combined with this text phrase that create six leading zeros)
I don't really know anything about hexadecimal so at this point I'm just taking shots in the dark and trying to guess number combos all night isn't really my thing. I don't necessarily need to solve this problem for anything besides practice but I am curious to know more about what's going on here. (And yes I am aware that MD5 is compromised and I wouldn't ever use it in production)
This problem can only be solved by brute-forcing. That is exactly how "proof-of-work" in Bitcoin works, for example. The only way to speed it up is to optimize each step in your calculation. Bitcoin miners have moved to specialized hardware because of this. They don't do anything "special" or "clever", they just calculate hashes very, very fast.
You can only optimize the code and throw more/better hardware at it. A cluster of compute nodes would also work well here, the problem lends itself to parallel processing (again, Bitcoin mining pools).
If you have a multi-core CPU, an easy thing to do is to use one Thread per CPU. Should speed up linearly (which may still not be fast enough).
I'm also working through the Advent of Code but I'm using PowerShell. I solved Puzzle 2 of Day 4 with basically the same code that I used to solve Puzzle 1. The only change was to the WHILE condition to check for six leading zeroes. It did take a lot longer to run than it did to solve for Puzzle 1. I just kicked it off and went to bed after I saw this post. Got my answer when I woke up. My code is posted at Github. Advent of Code Day 4 Puzzle 2 solution with PowerShell
I was searching trough google for an answer but i can't find something usefull. I am making games with libGDX and till now performance was not a big issue. But the game that i am working on now, will need some more optimization so here is my question:
With libGDX there is a lot of floats. I know that int is faster but what if i cast floats into int? Is this faster then using a float numbers or should just go with floats?
I don't need very high precision, becouse i mostly use that numbers for coordinates. But with much multiplications (i don't use division becouse it is slower) though the code i wonder what i should use.
Without even seeing your code the answer is: you're probably looking in the wrong place. The performance problem is not due to the use of floats, and switching to ints will not fix it.
As soon as you get performance problems (and even before) you must find ways to benchmark and measure the performance of your code. You might use profiling, or internal tracing, or some other way but you must do it.
Most performance problems are due to bad algorithms and/or a few small functions called millions of times. Fix the algorithms and recode the small critical functions and you will probably solve the problem. Switching your code to int is something you do after you've fixed everything else and need just a little more.
Put up some code on CodeReview and I'm sure you'll get lots more help with it.
I am working on a project in Android for my Signal Processing course. My aim is to find signal properties, such as periodicity, even/odd, causality etc, given a user-input function. Right now, I am stuck at trying to figure out how to programmatically calculate the periodicity of a given function. I know the basics behind periodicity: f(t+T) = f(t)
The only idea I have right now is to extensively calculate values for the function, and then check for repetition of values. I know the method is quite stupid, given the fact I don't know how many such values I need to calculate to determine if it is periodic or not.
I know this can be done easily in Matlab, but again very difficult to port Matlab to Java. Is there something I am missing? I have been searching a lot, but haven't found anything useful.
Thanks for any help, in advance!
If the function f is given as a symbolic expression, then the equation you stated can in some cases be solved symbolically. The amount of work required for this will depend on how your functions are described, what kinds of functions you allow, what libraries you use and so on.
If your only interaction with the function is evaluating it, i.e. if you treat the function description as a black box or obtain its values from some sensor, then your best bet would be a Fourier transformation of the data, to convert it from the time domain into frequency domain. In particularly, you probably want to choose your number of of samples to analyze as a power of two, and then use FFT to quickly obtain intensities for various frequencies.
I know there are other posts about this, but I cannot seem to find one strictly for handwriting. I am going to have a form and all I need to read in is 8 squares in the left hand corner that will have 3 letters proceeded by 5 numbers.
The problem with most posts is that people either post about software for writing on the screen or software that doesn't recognize handwriting yet. I would prefer to have something in java, but something simple in another language would work.
What would really work is if people could scan their documents and just type the job number for the document name, but apparently they cant do that right...
Can you change the form? This problem will simplify a lot if you can change the form to be something that is easier for a machine to read. To recognize an arbitrary handwriting is hard as well as error prone.
What I have in mind is a form like this:
form example http://shareworldonline.com/w3/testprep/images/test%20form.jpg
However, if you have to have handwriting, check out the solutions in this thread.
if i got you correctly, you are doing offline hwr,
when i was doing offline hwr, i found most difficult separating characters in word, seems like you have them in squares, so all what you need to do is find your boxes (ie by using histogram)
and compare content of your box with each element in yours characters database (i used levenshtein distance for that)
I know it maybe not very helpful, but maybe push you on right track.
This question already has an answer here:
Closed 13 years ago.
Possible Duplicate:
Need help solving Project Euler problem 200
Similar to this question
Project Euler Problem 200.
I wrote up a brute force solution in Java that takes several hours to run, and produced the first 500+ sqube numbers, which I thought should be enough. However, none of the answers from 190 to 210 seems to be the correct answer.
I'm wondering what I'm doing wrong here and how I could optimize this. Could the problem lie in BigInteger.isProbablePrime()?
I'm not sure if Stackoverflow is the best place to ask this, but I seem to be stuck. I've included my code and the generated data.
I'd really appreciate it if someone would give me some hints or pointers.
Edit: I've run the program again simply with using the first 500,000 prime numbers; took a day to run but produced the correct answer.
I'm a Project Euler administrator. Please do not post information that can spoil the problem for others, particularly code and answers, even half-functioning code. Please edit your question accordingly. EDIT: Thank you for doing so!
It's not unusual for solvers to use the web to search for information on solving a problem, and it would take away some fun if they stumbled upon such a spoiler. (Yes, I know there are sites with lots of solutions ready-made, but at least they're generally for the lowered numbered easy problems.)
We have forums for discussing difficulties with problems and getting hints, which are aggressively edited for spoilers.
Aren't you supposed to think of a clever solution which doesn't take a day or even an hour to run? :D
I think the problem is isProbablePrime, which doesn't guarantee that a number is a prime. It just says that a prime which was found could be a prime with a certain probability.
You should use an algorithm which is certain that it has found a prime.
The first answer is incorrect because isProbablyPrime isn't always correct (hence the Probably). It's slow, in part, because you are using BigInteger. All the values involved will fit in a long. Why not use a long?
There may be some simple refactorings you could do that would save some time. It seems that the bulk of the time is spent in the nested for loop. The order of magnitude is n squared. You could reduce the complexity by not nesting loops. Another problem is that you are finding more potential results than are required. You only need to find 200. The reason you are needing to find more is due to the fact that you are not finding the potential results in their numeric order. The TreeSet does keep the results in order, but the algorithm would be faster if it were able to stop when the 200th result was found.