The docker refresh seems to have fixed the bug in "Sub-select Abundance Profile".  I suspect it was already checked into git and the source refresh did the trick (as well as added Roman's "Import Metagenome List" widget, so we have that now).  I wonder, is there a schedule of when the updates are happening so we can plan around that?  We shouldn't request one for every bug fix.

So, I made the workspace world readable.

  http://demo.kbase.us/narrative/ws.673.obj.3

Still having trouble with the metagenome metadata in terms of:

1) knowing what the fields and contents are
2) how to address the columns
3) an error with the "Statistical Significance Test" (at the bottom of the Narrative) which is probably choking on using metadata column 1 because I doubt there is a metadata column 1.

Please take a look and tell me what I'm doing wrong or if there's a bug fix that can be done and deployed rapidly.

Thanks,
- Dylan



On Fri, Feb 7, 2014 at 4:20 PM, Dylan Chivian <dcchivian@lbl.gov> wrote:
Of course.  Please send me a username to share it with.


If possible, can you tell me

* how to determine which metadata column numbers are relevant, how the metadata is stored and accessed

* whether you are also getting an error implementing a row cutoff in the "Sub-select  Abundance Profile" widget (see the bottom of the narrative) and if so, whether there's a fix that can be implemented rapidly so I can make screen caps today.


in function 'Sub-select Abundance Profile': invalid literal for int() with base 10: ''
Traceback (most recent call last):
1) in '_select_matrix' line 714: txt += " All but first %d rows were removed."%int(cols)

Thanks,
- Dylan



On Fri, Feb 7, 2014 at 3:54 PM, Folker Meyer <folker@anl.gov> wrote:
Dylan,

we have been working like crazy getting our narratives together and adapted to the current system. Nobody is ignoring you. My entire team has been burning the midnight oil plus worked several weekends for a while now.

Please do not assume anyone is ignoring you just because people don't sit here waiting for your email. Anyway right now most of the team is on their way home. We still have a ton of work to do.

If there are any problems it would help if people could see your narrative to provide suggestions.

Best,
Folker
PS: We have had someone Q&A the workflows/narratives and make sure the science story is ok for our workflows. Now we are focussed on the remaining technical issues.

On Feb 7, 2014, at 5:37 PM, Dylan Chivian <dcchivian@lbl.gov> wrote:

> Look, guys, I know you're busy, but I don't think responding to me can be that time consuming.  I've only had access to your Narrative for 3 days and have been working through the kinks (and hopefully helping you debug it).  We're almost done.  I (probably) just have the last two issues and then I can start making a final version I can use for screen shots.  Maybe it'll be less annoying if you think of me as performing a useful QA role?
>
> Thanks,
> - Dylan
>
>
>
> On Fri, Feb 7, 2014 at 2:49 PM, Dylan Chivian <dcchivian@lbl.gov> wrote:
> OK.  I've gotten the BoxPlot, Heatmap, and PCoA to work.  There are only 2 outstanding issues remaining at this point in me being able to implement your workflow with another data set.
>
> * I cannot tell which metadata field number corresponds to which metadata field when trying to use the PCoA widget.
>
> * There is an error when running "Sub-select Abundance Profile" when providing a value for the number of rows to capture.  It runs fine when there is no value provided for this argument.
>
>
> Thanks,
> - Dylan
>
>
>
>
>
> On Fri, Feb 7, 2014 at 12:27 PM, Folker Meyer <folker@anl.gov> wrote:
> Hi Dylan,
>
> sorry to hear about wife's surgery.
>
> People here are quite busy (and the crashed WS service does not help) finishing up the 3 narratives we have been working on.
>
> The error that you report does not show up for us (or did not when we still had WS services). So maybe you want to share you workspace with us.
>
> best,
> Folker
>
>
> On Feb 7, 2014, at 1:26 PM, Dylan Chivian <dcchivian@lbl.gov> wrote:
>
> > Hi Folker,
> >
> > I'm at the oral surgeon waiting for my wife right now, so I don't think I can do a call at noon.  I'll ping you when I get into the office, looking like about 1pm at the earliest, possibly a bit later.
> >
> > The good news is I got past the block point of creating Metagenome Collections (I just stuck the object into the workspace so I didn't have to wait any longer.  Roman has made a widget for this that we should add to the Communities widgets).  I have also succeeded in retrieving abundance profiles which can be viewed with the "View" widget.  I'm still waiting on a diagnosis from you guys on what's failing with the "2D Heatmap", "Box Plot", and "PCoA" plots before I can progress.
> >
> > Also, what's the trick for determining what the field names (and columns?) are in the metadata?
> >
> > Thanks,
> > - Dylan
> >
> >
> >
> > On Fri, Feb 7, 2014 at 10:57 AM, Folker Meyer <folker@anl.gov> wrote:
> > Hi Dylan,
> >
> > do we want to attempt another sync up call today? We can do it at 2pm our time, noon yours.
> >
> > Doing it early in the day gives us a chance to make modifications etc.
> >
> > It would be good to have a single slide deck going into next week.
> >
> >
> > best,
> > Folker
> >
>
>
>