Bokeh plot won't show minor ticks and ticks label on log scale overlaps with it self

Hello guys,

I’m trying to plot price on log scale and i can’t get the minor ticks to show on the y-axis. Also, the exponents on the log scale tick labels overlap with the base. I have tried lots of different workarounds found in the internet but apparently you can’t manually define minor ticks in bokeh and I couldn’t find a single topic talking about the self overlap on the labels. Code with random data to exemplify (some of the solutions i tried are in the code comments):

import pandas as pd
import numpy as np
from bokeh.transform import cumsum
from bokeh.models import HoverTool, Span, Label, ColumnDataSource, Line, Grid, LabelSet
from bokeh.io import output_notebook, show, output_file, curdoc
from bokeh.plotting import figure
from bokeh.layouts import gridplot
import bokeh.palettes as bplt

curdoc().theme = 'dark_minimal'
output_notebook()

def plot_price_comparison():
      
    # Data gathering and cleaning up
    
    x = list(range(0, 5000))
    df = pd.DataFrame({"days": x, "price": x * np.random.normal(loc=0, scale=90, size=5000)})
    
    # Setting figure up
    
    colors = bplt.all_palettes['Plasma'][4]
    p = figure(title = f"Price comparison"
               , x_axis_type = "linear", y_axis_type = "log"
               , x_axis_label = 'Days since bottom', y_axis_label = "ROI"
               , plot_width = 1200, plot_height = 800
               , background_fill_color = "#000000", outline_line_color = "white")
    
#     p.xaxis.ticker = list(range(0, 2200, 200))
    p.xaxis.major_tick_line_color = "white"
    p.xgrid.grid_line_color = 'white'
    p.xgrid.grid_line_alpha = 0.5
#     p.xaxis.ticker.num_minor_ticks = 10
    
#     p.yaxis.ticker = np.logspace(0, 3, num=20, endpoint=True)
#     p.yaxis.ticker = [1, 10, 100, 1000]
    p.yaxis.axis_label_standoff = 20
#     p.yaxis.major_tick_line_color = "white"
#     p.yaxis.minor_tick_in = 10
    p.ygrid.grid_line_color = 'gray'
    p.ygrid.grid_line_alpha = 0.5
#     p.yaxis.major_label_overrides = {1: '1', 10: '10', 100: '100', 1000: "1000"}

    p.min_border_left = 120
    p.border_fill_color = "black"
    
    # Plotting data
        
    source = ColumnDataSource(df)
    line = p.line(x = "days", y = "price", legend_label=f"Price",
           line_color=colors[0], line_width=4, source=source)
    h = HoverTool(tooltips=[("return", "@return"),], mode = 'vline', renderers = [line])
    p.add_tools(h)
    
    show(p)
    return df

plot_price_comparison()

Resulting figure:

It’s not possible to speculate without version information.

Running BokehJS 2.3.2 and Python 3.8.10. (Forgot to say!)

There were some formatting issues fixed in 2.3.3 so my first suggestion is to try upgrading to see if that solves the problem. If not, then it is just a bug and a GitHub Issue with full details would be appropriate.

Updated bokeh, still didn’t work as intended. I wil post a github issue.