Thanks, Bradley. I really like your example and in fact I have played with 
pmap already. I think it is a great tool for getting into distributed 
computing since - as far as I know - pmap sends the different input 
variables to different workers and communicates back the result). 

In some cases shared memory access might be more feasible (such as in the 
example I posted above). Does anybody how to do that in parallel?



On Tuesday, September 9, 2014 3:42:02 PM UTC-4, Alex wrote:
>
> Bradley, 
>
> That's an awesome tutorial. Thanks for putting that together. 
>
>
> On Monday, August 18, 2014 7:32:17 AM UTC-7, Bradley Setzler wrote:
>>
>> I found that the easiest way was to use two files - one file contains the 
>> function to be run in parallel, the other file uses Require() to load the 
>> function in parallel, and pmap to call the function.
>>
>> I have a working example of the two-file approach here:
>>
>> http://juliaeconomics.com/2014/06/18/parallel-processing-in-julia-bootstrapping-the-mle/
>>
>> Best,
>> Bradley
>>
>>
>>
>>
>>
>> On Wednesday, November 6, 2013 10:08:38 PM UTC-6, Lars Ruthotto wrote:
>>>
>>> I am relatively new to Julia and doing some simple experiments. So far, 
>>> I am very impressed by it's nice and intuitive syntax and performance. Good 
>>> job!
>>>
>>> However, I have a simple question regarding parallel for loops the 
>>> manual could not answer for me. Say I am interested in parallelizing this 
>>> code
>>>
>>> a = zeros(100000)
>>> for i=1:100000
>>>   a[i] = i
>>> end
>>>
>>> In the manual it is said (and I verified) that 
>>>
>>> a = zeros(100000)
>>> @parallel for i=1:100000
>>>   a[i] = i
>>> end
>>>
>>> does not give the correct result. Unfortunately it does not say (or I 
>>> couldn't find it) how this can be done in Julia? Does anyone have an idea?
>>>
>>> Thanks!
>>> Lars
>>>
>>>

Reply via email to