y-values overlap when using rangebreaks

[u3Izx9ql7vW4]

picture of behavior: https://imgur.com/a/oSBgqvp

I believe the same issue is experienced by another user here: https://community.plotly.com/t/rangebreaks-with-bounds-and-values-show-overlapping-candlesticks/46067

code to reproduce this:
(example.pkl can be found on https://repl.it/join/pdmlylnv-angstscheisse )

import pandas as pd
import plotly.graph_objects as go

df = pd.read_pickle('example.pkl')

fig = go.Figure(data=[go.Candlestick(x=df['time'], open=df['open'], high=df['high'], low=df['low'], close=df['close'])])

fig.update_xaxes(
    rangebreaks=[
        dict(bounds=[15, 9], pattern="hour"),
        # dict(bounds=["sat", "mon"]),
        dict(values=["2020-12-25"])
    ]
)
fig.show()

[fjafferi]

Any update on this? it seems the problem still exists.

[nicolaskruchten]

I'm having trouble loading the pickle file... can someone provide a simple reproducible example of this please?

[fjafferi]

AAPL_1H.zip
Here is a pickle file, I just tried with the code above and it seems okay. However, i was seeing the described behavior above yesterday. Let me reproduce and provide an example.

[fjafferi]

`import pandas as pd
import plotly.graph_objects as go

df = pd.read_pickle('AAPL_1H.pkl')

fig = go.Figure(data=[go.Candlestick(x=df.index, open=df['open'], high=df['high'], low=df['low'], close=df['close'])])

fig.update_xaxes(
rangebreaks=[
dict(bounds=[15, 9], pattern="hour"),
# dict(bounds=["sat", "mon"]),
dict(values=["2021-06-03"])
]
)
fig.show()`

So it seems the problem appears to be when passing the dates list, and if the date falls within the x range.
@nicolaskruchten

[nicolaskruchten]

@fjafferi Thanks for the zipfile but the problem is that I can't unpickle it :) Can you send me a CSV please?

[nicolaskruchten]

OK, I found the time to get this file parsed. With Plotly.py 5.1 (powered by Plotly.js 2.2) I'm not seeing this behaviour any more:

[nicolaskruchten]

Ah but when uncomment the weekend exclusion I do see a problem:

I've logged the issue in the underlying Javascript library as plotly/plotly.js#5783

[Lucasdesard]

I had the exact same problem and fixed it in the following way: I changed the order of the arguments that I passed to the function such as in the following example:
fig.update_xaxes(
rangebreaks=[
dict(values=["2022-12-25"]), # hide holidays (Christmas and New Year's, etc)
dict(bounds=["sat", "mon"]), # hide weekends, eg. hide sat to before mon
dict(bounds=[18, 6], pattern="hour") # hide hours outside of 6pm-6am
]
)

[jinseo99]

I am having similar issue as well.
For me the issue is when the holiday and weekend are in consecutive order.

There's two cases:

1. Sat, Sun, Mon (Holiday)
2. Fri (Holiday), Sat, Sun.

It seems the reordering of the rangebreaks arguments fixes them but only conditionally.

For example for case 1:
The following order works:

dict(values=["2023-05-29"]), # hide holiday that is Monday
dict(bounds=["sat", "mon"]), # hide weekends

and the following causes a bug and shows y-value overlaps:

dict(bounds=["sat", "mon"]), # hide weekends
dict(values=["2023-05-29"]), # hide holiday that is Monday

This occurs reversely for case 2:
The following order works:

dict(bounds=["sat", "mon"]), # hide weekends
dict(values=["2023-04-07"]), # hide holiday that is Friday

and the following causes a bug and shows y-value overlaps:

dict(values=["2023-04-07"]), # hide holiday that is Friday
dict(bounds=["sat", "mon"]), # hide weekends

[Coding-with-Adam]

hi @jinseo99
Thanks for reporting. This seems related to the plotly.js #5783 issue

I've updated the priority of that bug.