yerror bar in plot generated on vs code shows different level of thickness for yerror bar in comparision to saved pdf file.
In vs code :
In pdf format: (you can notice thickness has increased)
In png format (it looks fine in this format)
Please provide a MWE and information on the platform you are running on.
using VS code and Julia version 1.9.3
#MWE
using StatsPlots, CSV, Distributions, Measures, DataFrames
fake_norm_data = DataFrame(V = repeat(["A", "B", "C", "D", "E", "F", "G","H", "I"], inner = 51, outer = 1),
X = repeat(["a", "b", "c", "d", "d", "f", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s"], inner = 1, outer = 27),
Z = repeat(["2M", "3M", "5M"] , outer = 9, inner = 17),
Y = rand(0:0.1:2, 459),
W = rand(0:0.01:0.1, 459))
p1 = groupedbar(fake_norm_data.X[1:51], fake_norm_data.Y[1:51], group = fake_norm_data.Z[1:51], yerror = fake_norm_data.W[1:51], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - A", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p2 = groupedbar(fake_norm_data.X[52:102], fake_norm_data.Y[52:102], group = fake_norm_data.Z[52:102], yerror = fake_norm_data.W[52:102], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - B", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p3 = groupedbar(fake_norm_data.X[103:153], fake_norm_data.Y[103:153], group = fake_norm_data.Z[103:153], yerror = fake_norm_data.W[103:153], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - C", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p4 = groupedbar(fake_norm_data.X[154:204], fake_norm_data.Y[154:204], group = fake_norm_data.Z[154:204], yerror = fake_norm_data.W[154:204], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - D", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p5 = groupedbar(fake_norm_data.X[205:255], fake_norm_data.Y[205:255], group = fake_norm_data.Z[205:255], yerror = fake_norm_data.W[205:255], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - E", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p6 = groupedbar(fake_norm_data.X[256:306], fake_norm_data.Y[256:306], group = fake_norm_data.Z[256:306], yerror = fake_norm_data.W[256:306], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - F", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p7 = groupedbar(fake_norm_data.X[307:357], fake_norm_data.Y[307:357], group = fake_norm_data.Z[307:357], yerror = fake_norm_data.W[307:357], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - G", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p8 = groupedbar(fake_norm_data.X[358:408], fake_norm_data.Y[358:408], group = fake_norm_data.Z[358:408], yerror = fake_norm_data.W[358:408], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - H", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
p9 = groupedbar(fake_norm_data.X[409:459], fake_norm_data.Y[409:459], group = fake_norm_data.Z[409:459], yerror = fake_norm_data.W[409:459], msw = 0.6, alpha = 0.9, xrotation = 45, title = "Target - I", bottom_margin = 7mm, left_margin=5mm, legend = :outertopright)
plot( p1, p2, p3, p4, p5, p6, p7, p8, p9, layout = (3,3), size = (1600, 1200), left_margin=10mm, dpi = 300)
savefig("/path/dummy1_plot.png")
savefig("/path/dummy2_plot.pdf")
The default backend gr() doesn’t support the dpi attribute, so I am guessing it just ignores it and we get a weird looking plot.
Using pythonplot() as backend worked and I got:
NB, using inspectdr(), which also supports dpi, gets the error bars correctly, but messes up the grid and sizes:
For the record,
savefig("/path/dummy2_plot.svg")
and then manually to pdf worked fine as well (but I specified the dpi manually), so I am assuming the problem is in passing the pdf settings.
1 Like