I'm trying to understand what could cause this problem because everything seems to be fine but the result seems to point out some sort of buggy behavior.
I have a custom struct defined as:
struct MyStruct {
const u16 index;
...
MyStruct(u16 index) : index(index), ... { }
}
static const MyStruct array[] = {
MyStruct(__COUNTER__,...),
MyStruct(__COUNTER__,...),
MyStruct(__COUNTER__,...)
...
}
Now, if I check the preprocessed file with XCode everything is fine and I get incremental indices as they are supposed to be.
At runtime, instead, what happens is that the index 16 becomes 23 and the successive indices are lowered by one, eg:
real index stored index
0 0
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 23
17 16
18 17
19 18
20 19
This is quite curious, especially because 16 seems to be a specific bound. Compiler is Apple Clang 4.2 (based on LLVM 3.2), on XCode 4.6.
Related
Closed. This question needs debugging details. It is not currently accepting answers.
Edit the question to include desired behavior, a specific problem or error, and the shortest code necessary to reproduce the problem. This will help others answer the question.
Closed 6 years ago.
Improve this question
I have a for loop and a variable C.
The loop begins at 0 and is expected to run C times but actually it runs C-1 times.
Here is my code :
vector<int> v(1000) //Allocated globally
int M, S, C;
cin>>M>>S>>C;
//cout<<M<<" "<<S<<" "<<C;
int fi=0, la=0;
for(int i=0; i<C; i++)
{
int f;
cin>>f;
if(i==0l){ fi = f;}
v[f] = f;
cout<<i<<" "<<f<<" "<<v[f]<<endl;
if(i==C-1){ la = f;}
}
This is my test case -
3 27 16
2
3
5
6
8
9
10
13
14
15
16
19
20
21
22
27
Output by Xcode :
0 2 2
1 3 3
2 5 5
3 6 6
4 8 8
5 9 9
6 10 10
7 13 13
8 14 14
9 15 15
10 16 16
11 19 19
12 20 20
13 21 21
14 22 22
I use Xcode on Mac if it makes a difference.
The variables fi and la are to find the first and the last element of the list.
I want to know what is wrong in my code for the for loop and why is it not iterating C times.
Thanks
Your loop is iterating C times. This is the classic Zero-Based Numbering issue.
Let me explain using your example where C is 16 and a numbered list:
2
3
5
6
8
9
10
13
14
15
16
19
20
21
22
27
So you see your lopp did iterate 16 times. To go from 0 to 16 would have actually been iterating one more time, so 17 times.
Let's say I have these strings:
char ref[30] = "1234567891234567891";
char oth[30] = "1234567891234567891";
I want to use the SSE 4.2 _mm_cmpistri function in C++; Normally the string is parsed from left to right. Is there a way to tell the function to search in reverse (compare from right to left instead of left to right?
Instead of searching
--------------->
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
to search this way <-----------------
Later edit:
Here's what i want to do:
I have two strings and I need a function with this header:
int sse_cmp(const char *a, int posA, const char *b, int posB);
This function must compare the strings "backwords":
from posA to 0 or until posB == 0.
The function must return the number of of common chars from back.
Ex:
<--------- posA
a : 1 2 3 4 5 6 7 8 9
b : a b c d 7 8 9
<---- posB
will return 3 ( 987 )
What's the most efficient way to do it? ( with SSE )
You can use _SIDD_MOST_SIGNIFICANT as part of the mode parameter to _mm_cmpistri.
See Intel SSE4 programming reference
How to pick the best uniformed 1d array from the 2d arrays ?
I have two 2d array of : 11 x 10
Example :
4 8 12 12 12 14 16 18 4 1 0
5 7 11 12 13 11 15 18 3 2 1
8 3 12 14 18 19 20 21 8 5 4 ,
8 2 11 12 17 17 19 20 7 4 3 ,
4 7 11 11 11 15 17 19 5 1 1 ,
3 8 11 13 11 15 14 17 4 1 0 ,
4 7 12 13 13 14 16 19 3 1 1 ,
5 9 11 12 13 15 17 19 5 0 1 ,
9 7 25 22 24 18 23 17 3 3 3 ,
4 8 13 13 13 15 17 17 5 2 0 ,
here we have 2d arrays of size 11x10 - Need to analysis and have to find out the common 1d array which has common like.
find the best closing number and its difference- and keep doing for all the corresponding columns in an array .
below answer should be like - finding the corresponding very column and comparing with the next row column - if it has some difference below ( 5 ) take the both column of two rows are same and process for next column of the same row..process untill finding the 1 row where it has at least nearby matches of 5
4 8 11 12 13 13 15 18 4 1 0
why don't you do something like this
int[] count(int[][] array)
int result[11];
for(int x = o; x<= 11;x++)
{
int temp[10];
for(int y = o; y<= 10;y++)
{
temp[y] = array[y][x];
}
result[x] = getHighest(temp);
}
return result;
}
int getHighest(int[] array)
{
int size = array.length;
int[size][1] temp;
for(int x; x<= size;x++)
{
int z = array[x];
temp[z][0]++;
}
int highest = -1;
for(int z; z<= size;z++)
{
int z = array[x];
int h = temp[z][0];
if(h > highest)
{
highest = h;
}
}
return highest;
}
Something like this, but my C++ has gotten a bit of rusty so sorry if there are any mistakes.
I have a map. lets say map<int, vector<int> > mymap1.
I want to update mymap1 by deleting some “keys” and also removing unwanted “elements” from the vector part of the selected keys. The “key’ or the “element” going to be deleted is given from another vector, known as “mylabel”. Actually, What I need to remain in my map is the values whose label is equal to 1. (At the end, keys must have the elements whose label are 1 only.)
I have implemented this (see code below), but got some compiler errors.
map<int, vector<int> > mymap1;
map<int, vector<int> >::iterator map1;
for (map1=mymap1.begin();map1!=mymap1.end();map1++){
int key = map1->first;
if (mylabel[key].Label() != 1){ mymap1.erase(key);
}
else{
vector<int> &myvec = map1->second;
for (vector<int>::iterator rn=myvec.begin(); rn!=myvec.end(); rn++){
if (mylabel[*rn].Label() != 1) myvec.erase(myvec.begin()+(*rn));
}
}
}
for you to get an idea, i am showing some example of my map.
0 1 2 6 10
1 0 2 4 3 6
2 0 1 3 5 8
3 1 2 4 5 7
4 1 3 6 7
5 2 3 8 7 9
6 1 0 7 4
7 6 4 3 5 9 11 10 13 12
8 2 5 9 11 18 15 19 20 22
9 5 7 11 8
10 0 7 14 16
11 9 7 8 13
12 7 13 14
13 7 12 11 14 15
14 12 10 16 13 15 17
15 13 14 8 17 19
16 14 10 17 21
17 14 16 15 21 18
18 8 20 19 17 26 27
19 8 15 18
20 8 18
21 16 17 23 24
22 8
23 25 21 24 26
24 23 21
25 23 26
26 23 25 18
27 18 28
28 27
if i show you my mylabel, it is as follows.
for(int c=0;c<mylabel.size();c++){
cout<<c<<" : "<<"label "<<mylabel[c].Label()<<endl;
}
0 : label 0
1 : label 0
2 : label 0
3 : label 0
4 : label 0
5 : label 1
6 : label 0
7 : label 1
8 : label 0
9 : label 1
10 : label 0
11 : label 1
12 : label 0
13 : label 0
14 : label 1
15 : label 1
16 : label 1
17 : label 1
18 : label 0
19 : label 0
20 : label 0
21 : label 1
22 : label 0
23 : label 0
24 : label 0
25 : label 1
26 : label 1
27 : label 0
28 : label 0
When I am deactivating the else part and running above code I got an output. But, I want to say you that it is a wrong result. I am getting extra keys that should be deleted. I can’t figure out why I got this fault result.
if i show the list of keys what i got,
5
7
9
11
14
15
16
17
20 - wrong
21
24 - wrong
25
26
could you please help me to rectify my code in order to get my modified map. thanks in advance.
Your erasing logic is wrong, and you end up using invalid iterators. (You're literally pulling the rug out from under your feet if you erase an iterator and then keep using that iterator.)
For node-based containers (list, map, set, unordered), you typically erase as follows:
for (auto it = c.begin(); it != c.end(); )
{
if (must_delete(*it)) // or it->first
{
c.erase(it++); // advance first, then erase previous
}
else
{
++it;
}
}
(This patterns is my favourite justification for the post-fix increment operator.)
For contiguous containers (vector, deque), erasing one element at a time is inefficient, because it incurs repeated moves. The preferred idiom here is "remove/erase", but it requires that you supply a suitable predicate if you don't just want to remove straight by element value. Here's an example with lambdas, for brevity:
std::vector<int> v;
v.erase(std::remove_if(v.begin(), v.end(),
[](int n)->bool{return some_criterion(n);}),
v.end());
In your situation, you could write the lambda as [mylabel&](n)->bool{ return mylabel[n].Label() != 1; }; or write a traditional predicate object if you don't have lambdas:
struct LabelFinder
{
LabelFinder(const LabelVector & lv) : label(lv) { }
inline bool operator()(int n) const
{
return label[n].Label() != 1;
}
private:
const LabelVector & label;
};
Now use:
v.erase(std::remove_if(v.begin(), v.end(), LabelFinder(mylabel)), v.end());
The problem is in the for loop. std::vector<T>::erase() returns iterator to the new position followed by the erased item. So the loop should be written as:
for (vector<int>::iterator rn=myvec.begin(); rn!=myvec.end();)
{
if (mylabel[*rn].Label() != 1)
rn = myvec.erase(rn);
else
++rn;
}
Read the doc:
vector::erase()
By the way, I doubt on this:
rn = myvec.erase(myvec.begin()+(*rn));
Vs
rn = myvec.erase(rn);
Are you sure you want the first one?
An idiomatic way to erase elements which are not equal to one is this:
//Define this function
bool isNotOne(int n) { return n != 1; }
//then do this instead of writing manual loop
myvec.erase( remove_if(myvec.begin(), myvec.end(), isNotOne), myvec.end() );
It's called :
Erase-Remove Idiom
I am using Ubuntu 10.10, Codeblocks IDE, and gcc compiler. I noticed the program I am writing was creating some odd output. Eventually I narrowed the issue down to a for-loop in the program. I was surprised to discover that the following basic for-loop didn't perform as expected.
#include <iostream>
using namespace std;
int main()
{
for(unsigned int i = 0; i < 21; i++)
{
cout << i << endl;
}
return 0;
}
When I compile and run it, the output is:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Although one would expect the output should include zero. Very surprisingly, when I change the for loop to
#include <iostream>
using namespace std;
int main()
{
for(unsigned int i = 0; i < 20; i++)
{
cout << i << endl;
}
return 0;
}
I get the expected output of:
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
I can't for the life of me figure out why 21 (and all numbers greater than 21) give me this false output, while 20 (and lower numbers) don't. If anyone has run into anything like this before, I'd sure appreciate hearing how he/she worked around it.
maybe the screen just scroll?
try to redirect the output to a text file
This seemed so weird that i run your first program and got what i would expect :
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
However, i notice that you use gcc as your compiler. This one is aimed towards c programming. Better use g++ as i did for this. It works fine here. (i'm actually surprised gcc compiles that :/)