I have some code that I believe to run in O(n), however when I time it, it seems to run in polynomial time. I'm trying to process ~200,000 records, so I did it in blocks of size MAX_COUNT so I wouldn't run out of heap space. That is, during the processing phase, a few things take place that make the records increase dramatically in size.
I copied in the important parts from my code. I feel like something is going on here that has to do with ArrayLists that I just don't understand.
This might not be the smartest way to go about things, but I don't see why it's taking longer to process each block than the previous. That is, each bock is size 5000 (except the first block), but the 1st block processed takes ~5seconds, and the 20th block processed takes ~25seconds. I would expect them to all take the same amount of time.
// Maximum block size
final int MAX_COUNT = 5000;
// Total number of records in need of processing
int n = records.size();
// the number of blocks to process
int numBlocks = (n / MAX_COUNT) + 1;
if (n % MAX_COUNT == 0) numBlocks--;
// The number of records to process in the block.
int numRecords;
ArrayList<Record> recordBlock = null;
// Iterate backwards through the blocks.
for (int i = numBlocks; i > 0; i--) {
// Make sure we don't process too many records.
if ( (i == 1 && numBlocks = 1 && n % MAX_COUNT != 0) ||
(i == numBlocks && n % MAX_COUNT != 0) )
numRecords = n % MAX_COUNT;
else numRecords = MAX_COUNT;
recordBlock = new ArrayList<Record>();
//EDIT: Fixed loop syntax (typo!)
for (int j = numRecords -1; j >= 0; j--)
recordBlock.add(records.remove(j));
recordBlock = ThreadHelper.processRecords(recordBlock,true,true);
while (recordBlock.size() != 0) {
Record r = recordBlock.remove(recordBlock.size() -1);
// write 'r' to MySQL
}
}
This section
for (int j = numRecords -1; j >= j--)
recordBlock.add(records.remove(j));
will reallocate the backing array behind recordBlock every time the backing array is filled. Rather declare it as
recordBlock = new ArrayList<Record>(numRecords);
Also, the loop syntax is incorrect.
As already mentioned by #mcfinnigan
recordBlock = new ArrayList<Record>(numRecords);
In addition, replace
while (recordBlock.size() != 0) {
Record r = recordBlock.remove(recordBlock.size() -1);
// write 'r' to MySQL
}
by
for (Record r: recordBlock) {// write 'r' to MySQL }
recordBlock.clear();
There is a problem with the for loop adding to the recordBlock.
for (int j = numRecords -1; j >= j--)
recordBlock.add(records.remove(j));
should be
for (int j = numRecords -1; j >= 0; j--)
recordBlock.add(records.remove(j));
If I am not mistaken.
Edit:
Another mistake I found was in your if statement.
if ( (i == 1 && numBlocks = 1 && n % MAX_COUNT != 0) ||
(i == numBlocks && n % MAX_COUNT != 0) )
should be
if ( (i == 1 && numBlocks == 1 && n % MAX_COUNT != 0) ||
(i == numBlocks && n % MAX_COUNT != 0) )
Might I suggest simplifying it to:
if(i == numBlocks && n % MAX_COUNT != 0)
since the first condition is just a special case when i = 1.
if the real code has if and for statements without braces then what you think the code does may not actually be what it does.
Related
My code prints every number between two limits and makes a sum.
How can i print a new line every 10 numbers?
for (int i = nedreGrense; i <= øvreGrense; i++) {
sum = sum + i;
}
for (int tallStreng = nedreGrense; tallStreng < øvreGrense; tallStreng++){
System.out.print(tallStreng+"+");
}
System.out.print(øvreGrense+"="+sum);
Since your limits can be any integer, so if nedreGrense can be any integer,
then after nedreGrense increases by 10, its ones will not change only its tenth will change, to check we have to use % operator which gives ones of the number.
So use the condition if( (tallStreng != nedreGrense) && ((tallStreng - nedreGrense )%10 == 0))
for (int i = nedreGrense; i <= øvreGrense; i++)
sum = sum + i;
for (int tallStreng = nedreGrense; tallStreng < øvreGrense; tallStreng++){
if( (tallStreng != nedreGrense) && ((tallStreng - nedreGrense)%10 == 0))
System.out.println();
System.out.print(tallStreng+"+");
}
System.out.print(øvreGrense+"="+sum);
I'm looking for the equivalent word in the database by the ContextQuery method, and when a equivalent word is null the program must try to use the next index from words and add it up to the current to make it a two word, if the two word is still null the program will make it a three word looking for the next 2 value, the equivalent is now being printed in console but i have the error IndexOutOfBoundsExpection after running
for (int i = 0; i < words.size(); i++){
temp = QueryWithContext.query(words.get(i));
if((temp == null || temp.isEmpty()) && words.size() >= i+1)
{
QueryWithContext.query(words.get(i)+" "+words.get(i+1));
temp = QueryWithContext.query(words.get(i)+" "+words.get(i+1));
System.out.println("1st if");
if((temp == null || temp.isEmpty()) && words.size() >= i+2)
{
temp = QueryWithContext.query(words.get(i)+" "+words.get(i+1)+" "+words.get(i+2));
}
else
{
temp = words.get(i);
}
}
System.out.println(temp);
if((temp == null || temp.isEmpty()) && words.size() >= i+1)
must be
if((temp == null || temp.isEmpty()) && words.size() > i+1)
otherwise
words.get(i+1)
throws the IndexOutOfBoundsExpection.
The problem is most likely in this line: temp = QueryWithContext.query(words.get(i)+" "+words.get(i+1)+" "+words.get(i+2));. You are looping until i is less than the size of words, so i will range from 0 to n - 1.
The problem is that in your code, you keep going till i + 2 (and previously, i + 1). This is what is most likely causing your error. To fix this, see if you can do the following: for (int i = 0; i < (words.size() - 2); i++){
Alternatively, do as #Uli recommends.
I want to loop infinitely using a for loop if a number equals 0, and loop until that number number if the number is greater than 0. Here's the code to help visual what I'm getting at.
for (int i = 0; i < this.getNumRounds(); i++) {
// 30 some lines of code
}
or
for ( ; ; ) {
// 30 some lines of code
}
if getNumRounds() is greater than 0, do the first loop, if it equals 0, do the second. I would prefer to do this without copying and pasting my 30 some lines of code twice and using an if statement seeing as the code is redundant, though I could use a function to take out that redundancy, but I'm looking to see if there's another option.
Use the powerful ternary operator:
for (int i = 0; this.getNumRounds() == 0 ? true : i < this.getNumRounds(); i++) {
// 30 some lines of code
}
As noted in the comments by yshavit, there is a shorter, cleaner way of expressing this:
for (int i = 0; this.getNumRounds() == 0 || i < this.getNumRounds(); i++) {
// 30 some lines of code
}
Have you thought about using a while loop instead?
int i = 0;
while(i < this.getNumRounds() || this.getNumRounds() == 0) {
//some 30 lines code
i++
}
So you want something like this:
int num = //whatever your number equals
if (num == 0) {
boolean running = true;
while (running) {
doLoop();
}
} else {
for (int i = 0; i < num; i++) {
doLoop();
}
}
private void doLoop() {
//some 30 lines of code
}
This code puts the contents of the loop in a separate method and checks if the number is equal to 0. If it is, the program runs the doLoop() method forever. Otherwise, it runs until i equals the number.
While it would be better to just create a method and use an if-statement you could add an if statement inside the for-loop to decrease i every iteration. It would look like:
for (int i = 0; i <= this.getNumRounds(); i++) {
if(this.getNumRounds() == 0){
i--;
}
// 30 some lines of code
}
Notice I changed i < this.getNumRounds() to i <= this.getNumRounds. This way if the number of rounds is zero then the loop will be called.
You could do the following.
for (int i = 0; i < this.getNumRounds() || i == 0; ++i) {
do {
// 30 lines of code
} while (this.getNumRounds() == 0);
}
If getNumRounds is non-trivial to compute, consider pulling it out of the loop and calling it only once.
I am exceptionally new to programming, but I am working on improving my skills as a programmer. Recently, I gave myself the challenge to determine what multiples of a given number are made up of distinct digits. I have gotten most of it to work, but I still need to make the code apply for every number that is a multiple of the input one. The code I have working so far is as follows:
Integer numberA = 432143;
Integer numberB = numberA;
Integer[] digitArray = new Integer[numberA.toString().length()];
int index;
for (index = 0; index < digitArray.length; index++) {
digitArray[index] = (numberA % 10);
numberA /= 10;
}
int repeats = 0;
for (int i = 0; i < digitArray.length; i++) {
for (int j = 0; j < digitArray.length; j++) {
if ((i != j) && (digitArray[i]==digitArray[j])) repeats = repeats + 1;
}
}
if (repeats == 0) {
System.out.println(numberB);
}
This will determine if the number is made up of distinct digits, and, if it is, print it out. I have spent quite a bit of time trying to make the rest of the code work, and this is what I've come up with:
Integer number = 1953824;
Integer numberA = number;
Integer numberB = numberA;
for (Integer numberC = number; numberC.toString().length() < 11;
numberC = numberC + number) {
Integer[] digitArray = new Integer[numberA.toString().length()];
int index;
for (index = 0; index < digitArray.length; index++) {
digitArray[index] = (numberA % 10);
numberA /= 10;
}
int repeats = 0;
for (int i = 0; i < digitArray.length; i++) {
for (int j = 0; j < digitArray.length; j++) {
if ((i != j) && (digitArray[i]==digitArray[j])) repeats = repeats + 1;
}
}
if (repeats == 0) {
System.out.println(numberB);
}
}
I can't figure out why, but his just prints whatever the number is a bunch of times if it is made up of distinct digits, and leaves it blank if it is not. If anyone could tell me why this is occurring, or even tell me what I need to do to fix it, that would be superb. Remember, I am very new to programming, so please give a short explanation for any terms you use that are at all out of the ordinary. I am eager to learn, but I currently know very little. Thank you for your time, and I greatly appreciate any and all help you can give me.
You assign the value of numberA to numberB (which is the value of number) right before the for loop. After that, numberB is never modified or assigned to a new value, so for every pass through the for loop, you're simply printing the value of numberB, which is always 1953824 in this case.
There are several corrections that can be made to achieve the result you desire, while cleaning up the code a little. The first thing is to change the print statement to print the correct number:
System.out.println(numberC);
Since numberC is the variable that is being updated by the for loop, that's what you'll want to conditionally print out if there are no repeat digits. Since we've replaced numberB with numberC, that means numberB is not longer needed, you can delete the declaration for it.
Now, the next issue is when you're defining the digital array - you should use the length of numberC, not numberA. Also, inside the for loop, you should assign numberA the value of numberC, or else eventually nothing but 0s will be stored in your digitArray. Overall, here's what it should look like.
Integer number = 1953824;
Integer numberA = number;
for (Integer numberC = number; numberC.toString().length() < 11;
numberC = numberC + number) {
Integer[] digitArray = new Integer[numberC.toString().length()];
numberA = numberC;
int index;
for (index = 0; index < digitArray.length; index++) {
digitArray[index] = (numberA % 10);
numberA /= 10;
}
int repeats = 0;
for (int i = 0; i < digitArray.length; i++) {
for (int j = 0; j < digitArray.length; j++) {
if ((i != j) && (digitArray[i] == digitArray[j]))
repeats = repeats + 1;
}
}
if (repeats == 0) {
System.out.println(numberC);
}
}
This should produce the desired result. It seems to work on my machine :)
If you want, you can take Jeffrey's suggestion and change the Integer to the primitive int to avoid the overhead of boxing. However, you still need to use the Integer class to use the toString() method, but you can accomplish that using Integer.valueOf():
Integer.valueOf(numberC).toString()
So if I understand correctly you are trying to find out if certain multiple exists within your number. Instead of constantly dividing by 10 instead use the modulus symbol. You can even embed it in conditional statements.
For example:
if(numberOne % 2 == 0)
Then we know that numberOne divided by 2 has a remainder of zero and is thus a multiple of 2
What is the best way to check if a string index is in bounds? Let's say we are checking a String for index i-1 or i+1 because you cannot say != null.
Example:
for (int i = 0; i < string.length(); i++)
{
if (string.charAt(i+1) == '#' && string.charAt(i - 1) != '1')
{
}
}
Should you just check the length of the string and see if it is within it?
string.charAt(i+1) == '#'
Yes, I think you need to make sure i+1 is not greater than String length.
Example:
if( (i+1) < string.length() && (i-1) >= 0 && (yourcode))
{
}
Why not just check the length of the string?
if(myString.length() - 1 > i)
I've always modified the length (and/or start) of the loop in these cases... these are all good answers - there's no one way to do it, but this is how I would do it:
for (int i = 1; i < string.length() - 1; i++)
{
if (string.charAt(i+1) == '#' && (i != 1 || string.charAt(i - 1) != '1'))
{
}
}
EDIT: After considering all of the situations that this code entails, I wouldn't necessarily do this in this manner, but find a cleaner way to express exactly what I'm trying to accomplish
The simplest way is to change your for loop like this:
for (int i = 1; i < string.length() - 1; i++)
{
if (string.charAt(i+1) == '#' && string.charAt(i - 1) != '1')
{
}
}
Explanation:
string.charAt(i - 1) implies that you might read position i-1 of the string. Since 0 is the lowest valid value you can start the iteration with i = 1
string.charAt(i + 1) implies that you might read position i+1 of the string. Since string.length() - 1 is the highest valid value you need to end your iteration at i == string.length() - 2. Since the for-loop checks the condition before entering the loop using i < string.length() - 1 will be just right.
I would try with this code.
for (int i = 1; string!= null && string.length() >= 3 && i < string.length() - 1; i++)
//it solve the case of String == null and string too short
//it optimize the max number of cycles
{
if (string.charAt(i+1) == '#' && string.charAt(i - 1) != '1')
{
}
}