Boxplot (Javascript) change x,y axis scale

KNIME Analytics Platform
1

Good evening,

I am using the Node Boxplot (JavaScript) to plot different boxplots where I want to change the scale of the axis. I read something about writing CSS code in Generic Javascript View and import it as a flow variable to change the x and y axis.
Since I am not familiar with CSS Code, could someone please explain how to write the CSS code to change the scale of the x and y axis? It would be a great help to me!

Thanks in advance
Jasmin

2

Hi Jasmin,
unfortunately the css only works for styling the view or hiding certain elements.

Unless there is an hack I am not aware off,
you cannot change the the axis scale just by using the new css node.

You could probably change the size and presence of ticks and labels,
but the visualization will stay exactly the same,
not affected at all by what you show or not show on the axis ticks.

Did you try to use the node Domain Calculator?
It automatically redefines the upper and lower bound
that are closer to the table minimum and maximum.
The range of the plot that defines the scale on the axis are the domain bounds in the table spec.
If you had the wrong bounds, you should see the visualization with less empty space
below and above the box plot.

You can also change the bounds with Edit Numeric Domain node,
anyway you get an error if you define bounds
that do not comprehend all the data points in the column.

Furthermore you could normalize with the normalizer node with min-max option.
This way different columns with different numerical domains will be similar
and the box plots will be scaled in the same range: [0 , 1]

attached a workflow you can take as example

box_plot.knwf (23.5 KB)

cheers
Paolo

2 Likes
3

Hi,

Same situation happens with node Line Plot(Java Script). I wish there is a feature to change the axis scale separately, or simply by entering the axis range manually. Because even after normalization, if the data gathers at one side, the problem remains.

Regards
Kefang

1 Like
4

Can you please take a screenshot of the problem on the line plot?
That way I could maybe find a workaround for you!

Cheers
Paolo