Gmail Calendar Documents Web Reader more »
Recently Visited Groups | Help | Sign in
Google Groups Home
performance question
There are currently too many topics in this group that display first. To make this topic appear first, remove this option from another topic.
There was an error processing your request. Please try again.
flag
  4 messages - Collapse all  -  Translate all to Translated (View all originals)
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
LAca  
View profile  
 More options Feb 17, 12:34 pm
Newsgroups: comp.lang.haskell
From: LAca <rizso...@gmail.com>
Date: Wed, 17 Feb 2010 09:34:51 -0800 (PST)
Local: Wed, Feb 17 2010 12:34 pm
Subject: performance question
hi All,

I'm trying to create a simple application, which parses CPP output
files and creates a graphviz dot file. You can find the source here:

  http://sites.google.com/site/rizsotto/daopp

I'm using test files ~25GB as input, the allocation is ~1.5GB. I tried
to profile the application, but found not too much. The (:) allocates
the most memory and consume the most processor time.

Can anyone give me a hint where to look at?

Thanks,
Laszlo


    Forward  
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Mark T. B. Carroll  
View profile  
 More options Feb 17, 12:58 pm
Newsgroups: comp.lang.haskell
From: "Mark T. B. Carroll" <Mark.Carr...@Aetion.com>
Date: Wed, 17 Feb 2010 12:58:24 -0500
Local: Wed, Feb 17 2010 12:58 pm
Subject: Re: performance question

LAca <rizso...@gmail.com> writes:
> Can anyone give me a hint where to look at?

My first instinct would be to look hard at the map operations you have:
I've had some laziness surprises from those in the past. I'm guessing
that the Map.insertWith only deals with small values anyway? Are you
sure that the foldWithKey is as strict as you hope? If anyone else
suggests anything, they're probably more likely to be correct than I am,
I've just learned to view Map stuff with suspicion whenever I'm tracking
down space leaks.

Mark


    Forward  
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Hans Aberg  
View profile  
 More options Feb 17, 3:10 pm
Newsgroups: comp.lang.haskell
From: Hans Aberg <haberg_20080...@math.su.se>
Date: Wed, 17 Feb 2010 21:10:35 +0100
Local: Wed, Feb 17 2010 3:10 pm
Subject: Re: performance question

LAca wrote:
> I'm trying to create a simple application, which parses CPP output
> files and creates a graphviz dot file. You can find the source here:

>   http://sites.google.com/site/rizsotto/daopp

> I'm using test files ~25GB as input, the allocation is ~1.5GB. I tried
> to profile the application, but found not too much. The (:) allocates
> the most memory and consume the most processor time.

> Can anyone give me a hint where to look at?

If you do random access, using arrays might improve performance
dramatically. Singly linked lists are really slow.

   Hans


    Forward  
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
LAca  
View profile  
 More options Mar 8, 4:30 am
Newsgroups: comp.lang.haskell
From: LAca <rizso...@gmail.com>
Date: Mon, 8 Mar 2010 01:30:33 -0800 (PST)
Local: Mon, Mar 8 2010 4:30 am
Subject: Re: performance question
thanks for the comments. just an update what i have tried and not
worked:

- put more `seq` wherever i can,
- create map for each file separately and merge them at the end.

but most surprisingly the small change, what made it:

- build the reverse graph (instead of which includes which, now i
collect which were included from where.)

it requires 10 times less memory. can't explain why does it help, but
it does! ;)

  Laszlo

On Feb 17, 9:10 pm, Hans Aberg <haberg_20080...@math.su.se> wrote:


    Forward  
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
End of messages
« Back to Discussions « Newer topic     Older topic »

Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
©2010 Google