Re: [h5md-user] proposal 100 - update

[Peter Colberg]

Hi Pierre,

On Wed, May 27, 2015 at 11:17:43AM +0200, Pierre de Buyl wrote:
> Here is an update on proposal 100 with code!
> 
> 1. Felix added "Proposition to mix scalar datasets and attributes". Using
> datasets actually makes more sense: else, one cannot store units in relation 
> to
> the time data. Using an offset attribute is more straightforward also than 
> step0
> and time0.

I also prefer the use of datasets with an optional offset attribute.

  \-- step[]
       +-- origin: Integer[]
  \-- time[]
       +-- origin: Float[]

step and time are HDF5 objects, so they can be copied or linked.

> 2. I propose to extend it by allowing
> 
> /step                    Dataset {4/Inf}
> /time                    Dataset {SCALAR}
> /value                   Dataset {4/Inf, 32, 3}

What does 4/ mean?

Both step and time should be scalar datasets, no?

> Then, time[i] = step[i]*time (I hope this is self-explaining, if not let me
> know).

I suggest keeping step and time independent:

step[i] = step_offset + i*step_interval
time[i] = time_offset + i*time_interval

Peter