Trim tooltip

[russHyde]

The text that is displayed when the user mouse-overs a node on the treeview contained lots of information.
Here, we trim down the text to just display the cluster ID for the hovered node.

The original code (in append_interactivity_data) made a lot of changes to the "$data" data.frame in a ggtree object (adding "defmuts", "allmuts", "mouseover", "colour_var" columns, and a column for each mutation regex that is passed into append_interactivity_data). Of these, the only columns that were actually used when setting up the interactive ggtree/girafe object were "colour_var" and "mouseover" and the mutation regex columns were used when making the genotype-heatmap.

We removed the "defmuts" column because this isn't presented by the tooltip, and isn't used in any other downstream processing step.

The "allmuts" column was not kept in either. But for different reasons. It is used to create the mutation-regex columns that the mutation heatmap is made from. Rather than defining these columns in append_interactivity_data (since they aren't presented 'interactively' to the user) they are now defined by extract_genotype_data where they are used to define the heatmap matrix and are thrown away.

[russHyde]

After the changes:

[russHyde]

Before the changes:

[russHyde]

I think the tooltip looks a bit scruffy. There's a preceding newline (possibly due to a hidden kable header) and there is the data.frame "Column.NAME" dot-separation mangling going on. I can fix these to just say

---------         
Cluster ID  NNNNN
---------

or just

Cluster ID: NNNNN

[jr-nicola]

LGTM - I'd probably go for the second option of having just:

Cluster ID: NNNNN

[russHyde]

Thanks @jr-nicola I'll modify the tooltip text

[russHyde]

Tooltip just displays "Cluster ID: NNNN" now:

[jr-nicola]

LGTM

[jr-nicola]

When I ran this updated version on the larger data set and tried to parse the tooltips to get the cluster ID as a numeric, I got a couple of examples where the label was:

Cluster ID: NA

I assume these were already in the previous version, and our parsing just didn't return an error the previous time?

[russHyde]

I've just reran the code here. Calling girafe(ggobj = tree, options = girafe_options) throws warnings (as it did above), but it no longer prints the girafe object to the Viewer panel in RStudio. Will investigate further.