problem with colorbar LaTeX title rendering

[remingtonsexton]

I am trying to implement a LaTeX math title for my colorbar and plotly is having some trouble rendering it. I am using Python 2.7, Plotly 4.5.2. My code:

fig = go.Figure()
fig.add_trace(go.Scatter(x=x_data, 
                         y=y_data,
                         mode='markers',text=obj_name,name='',
                         marker=dict(symbol='circle',size=12,line=dict(width=1,color='SlateGray'),
                                     opacity=0.75,color=log_mbh,colorscale='Rainbow',showscale=True,
                                     cmin=6.0,cmax=8.0,
                                     colorbar=dict(title=dict(text=r'$\Large\alpha\beta\gamma\delta$',side='top'),
                                                   thickness=50,tickmode='array',tickvals=[6.0,6.5,7.0,7.5,8.0]), 
                                    ),
                         )
             )
fig.add_trace(go.Scatter(x=np.arange(0,300,10), y=np.arange(0,300,10),
                         line = dict(color='LightSlateGray', width=2, dash='dash')))

fig.update_layout(
    template='plotly_white',
    xaxis = dict(showgrid=False,ticks='inside',zeroline=True,zerolinecolor='black',rangemode='tozero'),
    yaxis = dict(showgrid=False,ticks='inside',zeroline=True,zerolinecolor='black',rangemode='tozero'),
    xaxis_title=dict(text=r'$x$'),
    yaxis_title=dict(text=r'$y$'),
    font=dict(size=16),
    showlegend = False,
    autosize   = False,
    width      = 750,
    height     = 750,
    margin     = dict(l=50,r=50,b=50,t=50,pad=10),
)

fig.show()

Which produces the output:

As you can see, the LaTeX title seems to be rotated 90 deg, and ignoring side='top'. I have tried to increase the width of the colorbar and increase the plot's margins to no avail. Is there a workaround or solution for this problem?

[remingtonsexton]

Update: I have found a workaround but not a solution. This is definitely a bug as far as I cant tell.

One option, if you do not require Greek alphabet, is to simply use HTML code, simply using italics, superscripts or subscripts.

If one wants to use Greek letters, HTML entities don't work, but a few decimal and hexadecimal codes do work.

This issue should be addressed, especially if Plotly is to be used for publishing purposes. Otherwise, Matplotlib is the alternative that does work.

[theo-brown]

Any updates?

[Allyson-Robert]

At the risk of repeating the previous poster: Are there any updates on this or any way to work around the issue in another way? I use LaTeX extensively and it would require a complete rewrite to switch

[nicolaskruchten]

No updates or workarounds, sorry, this is just currently unsupported and as far as I know no one is working on this issue. The fix would need to be implemented in Plotly.js: plotly/plotly.js#4605

[Allyson-Robert]

No updates or workarounds, sorry, this is just currently unsupported and as far as I know no one is working on this issue. The fix would need to be implemented in Plotly.js: plotly/plotly.js#4605

Thanks for the reply. It really is unfortunate. I wish I had the skills to fix this myself

[empet]

To add a LaTeX title to colorbar I'm working with PlotlyJS.jl, a Julia lang version of plotly.
In Jupyter Notebook with Julia kernel all greek letters, subscripts, superscripts and more can be converted from Latex symbol to their unicode. As example, \xi followed by tab generates the corresponding letter ξ, then \_1 tab leads to ξ₁. Copy this string "ξ₁", and paste it after colorbar_title:

using PlotlyJS
fig = Plot(heatmap(z=rand(5, 5), colorbar_title="ξ₁/μ", colorbar_thickness=23),
           Layout(width=400, height=370))

Depending on your colorbar title, it is also possible to generate the unicode of a LateX symbol in a Jupyter Notebook with Python kernel. You can get all greek letters similarly: Latex and tab.
Another workaround is to search the web for the unicode of your LaTeX symbol.
For example replacing in both Julia and Python code, the above string assigned to colorbar_title, with colorbar_title="A\u2229B" we get:

[f0uriest]

Not an ideal fix, but converting the TeX to unicode with https://github.com/phfaist/pylatexenc worked for me.