Big variable names overlapping in output

[sillasgonzaga]

This is how big are the variable names I'm talking about:

> head(df2)
                                       sequencia volume
1 Critica Pedido-Alçada: CAF-Cód.Crítica: 000001     15
2 Critica Pedido-Alçada: CAF-Cód.Crítica: 000002     10
3 Critica Pedido-Alçada: CAF-Cód.Crítica: 000013      4
4 Critica Pedido-Alçada: CAF-Cód.Crítica: 000051      1
5 Critica Pedido-Alçada: CAF-Cód.Crítica: 000154     10
6 Critica Pedido-Alçada: ETP-Cód.Crítica: 000152     34
> tail(df2)
                                                                 sequencia volume
39 Devolução-Resp: 000008 - TRANSPORTADORA -Motivo: 000083 - PROPRIETARIO       7
40 Devolução-Resp: 000008 - TRANSPORTADORA -Motivo: 000087 - TEMPO INSUFIC     24
41 Devolução-Resp: 000008 - TRANSPORTADORA -Motivo: 000089 - PDV FECHADO A      4
42                                                            Falta Real--    120
43                                         Pedido Orfão-Orfão Faturamento-      4
44                                         Pedido Orfão-Orfão Roterização-      6

What happens is when I output this data.frame in sunburstR, the variable names overlap in the output and I can't see the percent result of that piece of the sequence, as I try to show in the image below:

http://imgur.com/LEeFZU8

Is there a workaround for this or should I use shorter strings for the variables?

[timelyportfolio]

I had not previously exposed the breadcrumb trail to customization. I'll look into it to see how easy it might be. I seem to remember this being a variable.

[timelyportfolio]

Thanks @sillasgonzaga, in 778cded I added an argument breadcrumb to customize the size of the breadcrumb. The issue though is the breadcrumb is set up as SVG and text-wrapping is difficult, so the trail will easily exceed the size of the page.

sunburst(sequences, breadcrumb = list(w = 200))

I will try to think of other strategies.

[timelyportfolio]

@sillasgonzaga I just added tasks functionality, which would allow styling with a smaller font-size.

sb <- sunburst(sequences,breadcrumb = list(w = 50))
sb$x$tasks <- htmlwidgets::JS(
  'function(){d3.select(this.el).select("#" + this.el.id + "-trail").style("font-size","60%")}'
)
sb

Again, not optimal but potentially helpful.

If you were ok with dumping the text and just using color in the breadcrumb, you could do something like this.

sb <- sunburst(sequences,breadcrumb = list(w = 10))
sb$x$tasks <- list(htmlwidgets::JS(
  'function(){d3.select(this.el).select("#" + this.el.id + "-trail").style("font-size","0")}'
))
sb

[sillasgonzaga]

Both the breadcrumb argument and the tasks functionality are wonderful additions to sunburstR. By using them and changing the parameters of sb$sizingPolicy$padding and sb$sizingPolicy$knitr$figure = FALSE I managed to arrange my messed dataset into an outstanding visualization in a knitr document.

Many thanks for the quick fix.

If you allow me to make you a suggestion, I would include the default values of the breadcrumb argument in the ?sunburst output

[timelyportfolio]

Will close for now? Feel free to reopen.

[timelyportfolio]

@sillasgonzaga pulls #13 and #14 should help with this if you want to give it a whirl. Would love testing.