Which one will be perform faster.

Question

1.44/5 (3 votes)

See more:

Hi Everybody,

There are two conditions which can provide me the output. But the issue is this which one is better and perform faster. Why? They both condition will provide the same output.

a b c d are string variables.

Case 1 :

if (a==c && b==d) {
   // Output
}

Case 2 :

if (a+ b == c+d) 
{
//Output
}

Posted 29-Oct-14 3:13am

Anubhava Dimri

Updated 29-Oct-14 3:30am

v2

Add a Solution

Comments

V. 29-Oct-14 9:17am

I would guess the first one will be faster, but you'll only know for sure when timing the two...

Anubhava Dimri 29-Oct-14 9:21am

Need reason....

ZurdoDev 29-Oct-14 10:15am

Then you run the tests and figure it out. Don't expect others to do your homework.

King Fisher 29-Oct-14 9:19am

Interview Question?

Anubhava Dimri 29-Oct-14 9:20am

no matter....

V. 29-Oct-14 9:29am

please be a little more polite. You're asking for help, not us, the least you can do is be a little courteous to those trying to help you...

KaushalJB 29-Oct-14 9:21am

Both would be same...as this notations are easy reading in runtime. But why this twist ?

Anubhava Dimri 29-Oct-14 9:29am

The reason behind this question is that "I need to know that "Is arithmetic operation is faster then Logical Operations"

PIEBALDconsult 29-Oct-14 10:15am

What difference does it make? And your example is probably not a good test.

BillWoodruff 29-Oct-14 9:23am

You are comparing very different things: what if a == 3, b == 7, c == 2, d == 8 : the first test yields false, the second test yields true.

Anubhava Dimri 29-Oct-14 9:30am

Suppose these variables are string

BillWoodruff 29-Oct-14 9:49am

Okay: that's a different question, entirely. See my comments in my response.

PIEBALDconsult 29-Oct-14 10:16am

The same problem still exists.

BillWoodruff 29-Oct-14 10:34am

string a = "dog"; string b = "food"; string c = "dog"; string d = "food";

(a==c && b==d) is always equal to (a + b == c+d)

how does "the same problem still exist" ?

PIEBALDconsult 29-Oct-14 10:36am

string a = "dog"; string b = "food"; string c = "dogf"; string d = "ood";

CPallini 29-Oct-14 10:39am

5!

BillWoodruff 29-Oct-14 11:04am

I just swallowed my dog-collar ! Can I plead "burn-out" ? Time to clock-out on QA for the day :)

PIEBALDconsult 29-Oct-14 11:09am

"Guess our minds must be too highly trained." :)

PIEBALDconsult 29-Oct-14 10:28am

Those two tests are not the same, so use whichever does what you need.
Generally, a logical should require less work, but it depends on how the test is implemented (perhaps in the chip).
Getting the right answer slowly is better than getting the wrong answer quickly.
I would definitely not do that with strings.

3 solutions

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

BillWoodruff · Accepted Answer · 2014-10-29T04:14:00

Whether a,b,c,d are integers (first version OP), or are strings (second version OP), comparing their possible speed of execution must take into account the distribution in the tests of values of 'a and 'b such that the 'a is not equal to 'c.

Because: the logical And operator && in C# will not execute the second condition (b == d) when (a != c); this is "short-circuit" optimization.

So, to the extent there are a large number of tests where a != c, the first test should be much faster than the second, which will always add 'a and 'b.

We'd also expect the first test to be faster in general because the addition of two strings creates a new string object, and that's done twice in the second test.

CPallini · Accepted Answer · 2014-10-29T04:46:00

Solution 3

No (as correctly noted by BillWoodruff and PIEBALDconsult) they don't provide the same output. You should focus your energy on understanding the difference of such pieces of code instead of worrying about performance.

Posted 29-Oct-14 4:46am

CPallini

Comments

Sergey Alexandrovich Kryukov 29-Oct-14 18:17pm

5ed!
—SA

CPallini 30-Oct-14 7:00am

Thank you, Sergey.

V. · Accepted Answer · 2014-10-29T03:27:00

Solution 1

C#

int a = 5;
int b = 7;
int c = 19;
int d = 7;

System.Diagnostics.Stopwatch sw = new System.Diagnostics.Stopwatch();

sw.Start();
for(int i = 0; i < 1000000; i++){
    if (a==c && b==d);
}
sw.Stop();

Console.WriteLine("(a==c && b==d) took " + sw.ElapsedMilliseconds + " milliseconds");

sw.Start();
for(int i = 0; i < 1000000; i++){
    if (a+b == c+d);
}
sw.Stop();
Console.WriteLine("(a+b == c+d) took " + sw.ElapsedMilliseconds + " milliseconds");
Console.ReadLine();

output:
(a==c && b==d) took 3 milliseconds
(a+b == c+d) took 7 milliseconds.

Since you so impolitely stated into the reply you need a reason (which you didn't state in the question) I will let you figure it out yourself.

[EDIT]You did state it, but still you could remain more polite[/EDIT]

Posted 29-Oct-14 3:27am

V.

Updated 29-Oct-14 3:36am

v2

Comments

KaushalJB 29-Oct-14 9:31am

Nice one

V. 29-Oct-14 9:34am

thanks :-)

Anubhava Dimri 29-Oct-14 9:37am

Thanks Great... Acceptable

BillWoodruff 29-Oct-14 9:54am

In "real-world" conditions it is unlikely that a,b,c,d are constants, which is what you are comparing here. Any test that attempts to be meaningful in terms of actual use in an application must vary the values of a,b,c,d.

V. 29-Oct-14 9:56am

I thought of that, but I figured in this example that would be overkill ;-)

PIEBALDconsult 29-Oct-14 10:17am

Be sure you don't compile that with optimizations (release mode).

Sergey Alexandrovich Kryukov 29-Oct-14 18:22pm

This is not enough. Besides, one may want to compare optimized code. This timing is incorrect in principle.
First of all, JIT compilation time should be excluded. This can be done by writing separate methods and timing them not on the first call, but after first call was previously done.
Besides, all timing tests should collect some statistics, because operational situation is random. Long loop is not a proper way of collecting statistics.
—SA

V. 31-Oct-14 2:52am

The long loop is to have a larger "difference". (milliseconds instead of micro or nanoseconds). And yes, you're right, you should rerun this thing over and over in a non-jit situation. But the principle is enough in this case. I also believe the outcome will still be the same, may it be with less difference. But we could test ;-)

Sergey Alexandrovich Kryukov 31-Oct-14 11:08am

No. You don't understand. You have to exclude JIT, but timing everything only when it was already called before. JIT compilation takes place on per-method basis. I don't understand that you are resisting. Fix the solution, that's it. Also, collect some more statistic, and you will see that long loops could not be enough; chanced are, you will observe considerable deviation...
—SA

V. 1-Nov-14 13:52pm

You are a REALLY annoying person. You´re not helping in any way and your only goal is to make sure everyone knows how clever you are. This is not the first time I feel this way about your replies. Please get off my back.

Sergey Alexandrovich Kryukov 1-Nov-14 17:27pm

Please calm down. Remember that you cannot be more free than other members. Let me decide what to comment and what not.
Also, please understand that we do not welcome personal accusations. If you don't like someone's comment, please address the problem under discussion, not a person.
—SA

V. 1-Nov-14 17:35pm

Let´s just drop the conversation. We both have opposite opinions...

Sergey Alexandrovich Kryukov 1-Nov-14 17:43pm

Sorry, no. I just want to help you to learn just another bit of timing technique, and you are resisting to learn it. If we really had opposite opinions, you would explain my mistake, but you did not try it.
No problem though. We can drop the conversation. I just would like to advise you not to cross the line.

Thank you for understanding.
—SA

V. 2-Nov-14 2:51am

Dude, It is YOU who crossed the line. I had and have no interest in going into depths of timing issues. If I ever had too I will be open for learning. If you have something to add to my solution, which I just conjured up in 10 minutes to give the guy an idea of how to do this sort of thing you could have created a solution 2 adding or improving your thoughts. Much more constructive without undermining anyone else. You´re smart enough, and it´s not about right or wrong. It´s about communication. You often cross this line (with everyone), which is why I reacted.

So you could say "thanks for understanding" in your previous reply and I can empathize you feel I don´t understand you, but from my point of view, you don´t (or didn´t) understand me.

With everything I have said up until now, you said I was wrong in some way. Something is not right there.

Sergey Alexandrovich Kryukov 2-Nov-14 15:55pm

I already added something to your solution, you just keep ignoring it...
Thank you for not understanding, or whatever do you prefer.
And please don't cross the line. All you write is visible to all member.
—SA

Sergey Alexandrovich Kryukov 29-Oct-14 18:20pm

There is on big fundamental flaw in this solution: you did not exclude JIT compilation time from your timing.
This is a dramatic and very usual mistake.
—SA

Which one will be perform faster.

3 solutions

Solution 2

Solution 3

Solution 1

Add your solution here

Preview 0

Existing Members

...or Join us