All colorblind friendly RColorBrewer palettes
recreated without the number of colors limitation
and with transparency support thanks to pal_alpha
that can be used alone.
Also, all viridis palettes (see the package on CRAN),
yet color ramps are borrowed and Momocs does not depend on it.
Also, pal_qual_solarized
based on Solarized: https://ethanschoonover.com/solarized/
and pal_seq_grey
only shades of grey from grey10
to grey90
.
pal_alpha(cols, transp = 0)pal_manual(cols, transp = 0)
pal_qual_solarized(n, transp = 0)
pal_seq_grey(n, transp = 0)
pal_div_BrBG(n, transp = 0)
pal_div_PiYG(n, transp = 0)
pal_div_PRGn(n, transp = 0)
pal_div_PuOr(n, transp = 0)
pal_div_RdBu(n, transp = 0)
pal_div_RdYlBu(n, transp = 0)
pal_qual_Dark2(n, transp = 0)
pal_qual_Paired(n, transp = 0)
pal_qual_Set2(n, transp = 0)
pal_seq_Blues(n, transp = 0)
pal_seq_BuGn(n, transp = 0)
pal_seq_BuPu(n, transp = 0)
pal_seq_GnBu(n, transp = 0)
pal_seq_Greens(n, transp = 0)
pal_seq_Greys(n, transp = 0)
pal_seq_Oranges(n, transp = 0)
pal_seq_OrRd(n, transp = 0)
pal_seq_PuBu(n, transp = 0)
pal_seq_PuBuGn(n, transp = 0)
pal_seq_PuRd(n, transp = 0)
pal_seq_Purples(n, transp = 0)
pal_seq_RdPu(n, transp = 0)
pal_seq_Reds(n, transp = 0)
pal_seq_YlGn(n, transp = 0)
pal_seq_YlGnBu(n, transp = 0)
pal_seq_YlOrBr(n, transp = 0)
pal_seq_YlOrRd(n, transp = 0)
pal_seq_magma(n, transp = 0)
pal_seq_inferno(n, transp = 0)
pal_seq_plasma(n, transp = 0)
pal_seq_viridis(n, transp = 0)
pal_qual(n, transp = 0)
pal_seq(n, transp = 0)
pal_div(n, transp = 0)
a palette function
color(s) as hexadecimal values
numeric
between 0 and 1 (0, eg opaque, by default)
numeric
number of colors
Default color palettes are currently:
pal_qual=pal_qual_Set2
pal_seq=pal_seq_viridis
pal_div=pal_div_RdBu
pal_div_BrBG(5) %>% barplot(rep(1, 5), col=.)
pal_div_BrBG(5, 0.5) %>% barplot(rep(1, 5), col=.)
Run the code above in your browser using DataLab