H5Part development and discussion

[Kurt Stockinger <KStockinger AT lbl.gov>]

Hi Thomas,

Thanks for your answers. Unfortunately I don't have access to your
machine. Can you put the file temporarily on a webserver so that I can
download it?

Kurt
> Hello Kurt,
>
> let me answer your comments/questions:
>
>   
>>> get file attribute info:
>>>
>>>   for (Int_t n = 0; n < fNFileAttr; n++) {
>>>     H5PartGetFileAttribInfo(fH5File,n,an,32,0,nElem);
>>>       
>> Is it intended that you don't return the type of the attribute (the 5th
>> parameter) or do you assume that you know the type, i.e. all are of type
>> char?
>>     
>
> I, uh, copied the command from somewhere else, and it has worked ever since.
> Now that I look at it, you are right, I should probably check for the
> type to make sure it is indeed a char...
>
>   
>>> where GetUnit(...) is a method that contains a loop over the file
>>> attributes to
>>> find the associated unit name:
>>>       
>> In the H5Part doc I didn't find anything specific about unit names. Can
>> you explain this a bit? Do you assume that the attribute name is unified
>> to char[32]?
>>     
>
> This is a "private" H5Part convention between Andreas and me: whenever an
> attribute string is followed by "Unit" it is understood to contain a string
> describing the unit of the attribute. Example: "x" -> "xUnit". In the 
> studies
> we are doing now, it is very important to know in what unit a given quantity
> is stored, in particular if you want to compare different files. If this 
> turns out to be a universally desirable feature, one could think about
> implementing it "officially" in H5Part.
> In H5root, I just use it to display the unit in square brackets on the 
> plots,
> there is no automatic unit conversion yet.
>
>   
>>> Now some numbers: a 3 Giga file with some 500 time steps (21 file
>>> attr., 12 step
>>> attr.) takes 19.1 s to read in (the second time only 0.5 s since the
>>> file is
>>> buffered). 
>>>       
>> This the time for reading the attributes only, right? If yes, then the
>> whole 3GB is not read entirely but only the 21*12 (char) attributes, right?
>>     
>
> See Achim's reply.
>
>   
>> Can you send me a pointer to the file to download so that we can look at
>> this together? This would also help me understand your performance a bit
>> better.
>>     
>
> The file is on merlin, /home2/adelmann/lcls-small.h5 . Does anyone have 
> access to merlin overthere?
>
> Cheers,
>
> Thomas
>   


-- 
Kurt Stockinger
Computational Research Division
Lawrence Berkeley National Laboratory
Mail Stop 50B-3238, 1 Cyclotron Road
Berkeley, California 94720, USA

Tel: +1 (510) 486 5208, Fax: +1 (510) 486 4004
email: KStockinger AT lbl.gov
http://sdm.lbl.gov/kurts/