[Opengeoscience-developers] Time-Series Visualization > Multiple Values for a location

[Ben Burnett]

On Thu, Oct 24, 2013 at 3:59 PM, darshan pandit <darshan12345 at gmail.com>wrote:

> I had a couple of concerns with regards to the time-series functionality
> I've been working on and would love your thoughts/inputs to go about this
> problem.
>
> 1. In the clt.nc, there is a plev dimension. As with the generic form, it
> is very highly likely that other datasets will also have 'level'
> dimensions. This results in multiple values for a variable even at a given
> time-point and a given location. We can render a visualization with
> time-series at mutiple-levels, even if the plevel values are continuous
> throughout and handsul (Something I have to check). But if there are too
> many level points or non-continuous plevels, we must think about a scheme
> to handle this.
>

You are right, there will definitely be datasets with additional
dimensions. For now, just select the first index in any extra dimension. In
the future we can make the function more generic and allow users to pass a
parameter for different dimensions, such as plevel.


>
> 2. The information is relatively sparse considering the fact that the
> values are floats and sending a very precise float value for location may
> not be always possible. for eg: 169.016 will give you data but the point at
> 169.017 may not. Is there any way to tackle this problem either from the
> cdms or the front-end during the selection.
>

Ideally we should compute the interpolated values, but just selecting the
nearest datapoint should be sufficient for now.


>  ,
>
> Regards,
> Darshan Pandit.
>
> _______________________________________________
> Opengeoscience-developers mailing list
> Opengeoscience-developers at public.kitware.com
>
> http://public.kitware.com/cgi-bin/mailman/listinfo/opengeoscience-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/opengeoscience-developers/attachments/20131104/65123e1f/attachment.html>