Wanted to add also examples for multi-band bars. However, so far I couldn't figure out how it works. Maybe @meghanrjones can help me out 😉.

Wanted to add also examples for multi-band bars. However, so far I couldn't figure out how it works. Maybe @meghanrjones can help me out 😉.

Is this what you're referring to?

import pygmt

data = {'x1': [0.5, 1.5], 'y': [1, 2], 'x2': [1.5, 0.5], 'x3': [0.75, 1.25], 'x4': [1.25, 0.75]}
df = pd.DataFrame(data=data)

fig = pygmt.Figure()
pygmt.makecpt(cmap="hawaii", series=[0,4,1])
with fig.subplot(
    nrows=1, ncols=2, subsize=("8c", "4c"), frame=["ag", "WSne"], sharey=True
):
    pen = "1.5p"
    with fig.set_panel(panel=0):
        fig.basemap(region=[0, 4, 0, 3], frame='+t"stacked bars"')
        fig.plot(data=data, style="B0.2i+i4", cmap=True, pen=pen)

    with fig.set_panel(panel=1):
        color = "tomato"
        fig.basemap(region=[0, 4, 0, 3], frame='+t"split bars"')
        fig.plot(data=data, style="B0.2i+i4+s4", cmap=True, pen=pen)

fig.show()

This example looks good to me 🙂!

May I ask a question regarding the general explanation at the top?

Here, +i means we must accumulate the bar values from the increments dy or dx, while +v means we get the complete values relative to base. Normally, the bands are plotted as sections of a final single bar. Use +s to instead split the bar into ny or nx side-by-side, individual and thinner bars.

What is the expected bar plot for only appending +v (no +s), maybe overlaid bars? I ask because the corresponding figure (lower right panel) appears not completely clear to me, especially the lower bar:
(Please note, this is definitely not meant in a way, that the example by @michaelgrund should be changed!)

import pygmt


# outline thickness of bars
pen = "1.5p"

# generate dictonary for plotting multi-band bars
data = {
    "x1": [0.25, 1.25],
    "y": [1, 2],
    "x2": [0.65, 0.5],
    "x3": [0.4, 1.25],
    "x4": [2.25, 0.75],
}

# set up colormap for multi-band bars
pygmt.makecpt(cmap="roma", series=[0, 4, 1])


# create figure object
fig = pygmt.Figure()

# make 2x2 subplot
with fig.subplot(
    nrows=2,
    ncols=2,
    subsize=("8c", "4c"),
    frame="ag",
    sharey=True,
    sharex=True,
    margins=["0.5c", "0.75c"],
    ):

    # top left: horizontal bars
    with fig.set_panel(panel=0):
        color = "tomato"
        fig.basemap(region=[0, 4, 0, 3], frame="+thorizontal bars (B)")
        fig.plot(x=1.5, y=0.5, style="B0.75c", color=color, pen=pen)
        fig.plot(x=3, y=1.5, style="B1c", color=color, pen=pen)
        # +b2 means that the bar is starting from x=2 here
        fig.plot(x=3.5, y=2.5, style="B0.5c+b2", color=color, pen=pen)

    # top right: grouped bars
    with fig.set_panel(panel=1):
        fig.basemap(region=[0, 4, 0, 3], frame="+tgrouped bars (+ v4 + s)")
        fig.plot(data=data, style="B0.75c+v4+s", cmap=True, pen=pen)
    
    # bottom left: stacked bars
    with fig.set_panel(panel=2):
        fig.basemap(region=[0, 4, 0, 3], frame="+tstacked bars (+ i4)")
        fig.plot(data=data, style="B0.75c+i4", cmap=True, pen=pen)
    
    # bottom right: overlaid ? bars
    with fig.set_panel(panel=3):
        fig.basemap(region=[0, 4, 0, 3], frame="+toverlaid ? bars (+ v4)")
        fig.plot(data=data, style="B0.75c+v4", cmap=True, pen=pen)

fig.show()
# fig.savefig(fname='question_bars_overlaid.png')

This example looks good to me slightly_smiling_face!

May I ask a question regarding the general explanation at the top?

Here, +i means we must accumulate the bar values from the increments dy or dx, while +v means we get the complete values relative to base. Normally, the bands are plotted as sections of a final single bar. Use +s to instead split the bar into ny or nx side-by-side, individual and thinner bars.

What is the expected bar plot for only appending +v (no +s), maybe overlaid bars? I ask because the corresponding figure (lower right panel) appears not completely clear to me, especially the lower bar: (Please note, this is definitely not meant in a way, that the example by @michaelgrund should be changed!)

Indeed, that's a good question at this point @yvonnefroehlich. Maybe @PaulWessel can help out again?

This example looks good to me 🙂!

May I ask a question regarding the general explanation at the top?

Here, +i means we must accumulate the bar values from the increments dy or dx, while +v means we get the complete values relative to base. Normally, the bands are plotted as sections of a final single bar. Use +s to instead split the bar into ny or nx side-by-side, individual and thinner bars.

What is the expected bar plot for only appending +v (no +s), maybe overlaid bars? I ask because the corresponding figure (lower right panel) appears not completely clear to me, especially the lower bar: (Please note, this is definitely not meant in a way, that the example by @michaelgrund should be changed!)

import pygmt


# outline thickness of bars
pen = "1.5p"

# generate dictonary for plotting multi-band bars
data = {
    "x1": [0.25, 1.25],
    "y": [1, 2],
    "x2": [0.65, 0.5],
    "x3": [0.4, 1.25],
    "x4": [2.25, 0.75],
}

# set up colormap for multi-band bars
pygmt.makecpt(cmap="roma", series=[0, 4, 1])


# create figure object
fig = pygmt.Figure()

# make 2x2 subplot
with fig.subplot(
    nrows=2,
    ncols=2,
    subsize=("8c", "4c"),
    frame="ag",
    sharey=True,
    sharex=True,
    margins=["0.5c", "0.75c"],
    ):

    # top left: horizontal bars
    with fig.set_panel(panel=0):
        color = "tomato"
        fig.basemap(region=[0, 4, 0, 3], frame="+thorizontal bars (B)")
        fig.plot(x=1.5, y=0.5, style="B0.75c", color=color, pen=pen)
        fig.plot(x=3, y=1.5, style="B1c", color=color, pen=pen)
        # +b2 means that the bar is starting from x=2 here
        fig.plot(x=3.5, y=2.5, style="B0.5c+b2", color=color, pen=pen)

    # top right: grouped bars
    with fig.set_panel(panel=1):
        fig.basemap(region=[0, 4, 0, 3], frame="+tgrouped bars (+ v4 + s)")
        fig.plot(data=data, style="B0.75c+v4+s", cmap=True, pen=pen)
    
    # bottom left: stacked bars
    with fig.set_panel(panel=2):
        fig.basemap(region=[0, 4, 0, 3], frame="+tstacked bars (+ i4)")
        fig.plot(data=data, style="B0.75c+i4", cmap=True, pen=pen)
    
    # bottom right: overlaid ? bars
    with fig.set_panel(panel=3):
        fig.basemap(region=[0, 4, 0, 3], frame="+toverlaid ? bars (+ v4)")
        fig.plot(data=data, style="B0.75c+v4", cmap=True, pen=pen)

fig.show()
# fig.savefig(fname='question_bars_overlaid.png')

Looks like an upstream bug.

Looks like an upstream bug.

Meaning I should open an issue in the upstream GMT repository?

Looks like an upstream bug.

Meaning I should open an issue in the upstream GMT repository?

Could you please make a minimal CLI example to reproduce the issue, and post it in this PR. Then we can ping Paul to let him see if it's a bug.

Output of gmt --version: 6.4.0

Data: data_vhbars.txt

Colormap (please remove the .txt): roma_vhbars.cpt.txt

CLI minimal example:

gmt plot data_vhbars.txt -R0/4/0/3 -JX8c/4c -SB0.75+v4 -Croma_vhbars.cpt -Ba1f1g1 -W1p,black -png vhbars_overlaid

Output figure:

Output of gmt --version: 6.4.0

Data: data_vhbars.txt

Colormap (please remove the .txt): roma_vhbars.cpt.txt

CLI minimal example:

gmt plot data_vhbars.txt -R0/4/0/3 -JX8c/4c -SB0.75+v4 -Croma_vhbars.cpt -Ba1f1g1 -W1p,black -png vhbars_overlaid

Output figure:

Ping @PaulWessel to see if this example makes sense to you.

@yvonnefroehlich Could you please open a separate issue in the GMT repository instead so that we can get some feedback from Paul or others?

Besides the potential upstream bug, I think this PR is ready for merge.

@yvonnefroehlich Could you please open a separate issue in the GMT repository instead so that we can get some feedback from Paul or others?

@seisman: I opened a related issue in the upstream GMT repository.

Besides the potential upstream bug, I think this PR is ready for merge.

@michaelgrund: I apologize for the delay in getting this PR merged caused by my question!