[BUG]The plot disappears when I click on a Rank legend for POSIX data

[aparajit7]

Describe the bug 🖍️

The entire POSIX plot disappears when I click on an individual rank to visualize the data.

Reproducing the bug 🔍

   import plotly.graph_objects as go
   from dash import Dash, html, dcc
   from plotly_resampler import FigureResampler, register_plotly_resampler
   import pandas as pd
   import numpy as np
   import time


   df = pd.read_csv('sample-enlarged.csv')


  df['mark'] = (df['start'] + df['end']) / 2
  df['mark_size'] = (df['end'] - df['start']) / 2
  df['rank'] = df['rank'].astype("category")
  df['color'] = '#000000'
  df.loc[df['operation'] == 'write', 'color'] = 'blue'
  df.loc[df['operation'] == 'read', 'color'] = 'red'

  df = df.sort_values(by=['mark'])
  df['mark'] = df['mark'].interpolate()

  df = df[df['api'] == 'POSIX']

  register_plotly_resampler(
     mode="LTTB",  
     default_n_shown_samples=500,  
  )

  fig = FigureResampler(go.Figure())
  start_time = time.time()

  for rank in df["rank"].unique():
     df_rank = df[df["rank"] == rank]
    
    
     df_sample = df_rank.sample(100)

     fig.add_trace(
         go.Scattergl(
             x=df_sample["mark"],
             y=df_sample["rank"],
             name=f'Rank {rank}',
             marker=dict(
                 color=df_sample["color"].iloc[0], 
             ),
             line=dict(
                 color=df_sample["color"].iloc[0],  
             ),
             error_x=dict(
                 type='data',  
                 symmetric=True,
                 array=df_sample["mark_size"],
                 color=df_sample["color"].iloc[0],
                 visible=True,
                 thickness=6,
                 width=0
             ),
             showlegend=True
         ),
         hf_x=df_rank["mark"],  
         hf_y=df_rank["rank"]
     )

  end_time = time.time()
  print(f"Script runtime: {end_time - start_time:.2f} seconds")

  app = Dash(__name__)

  app.layout = html.Div(children=[dcc.Graph(id="graph-id", figure=fig)])


  fig.register_update_graph_callback(app, "graph-id")

  if __name__ == '__main__':
      app.run(debug=True)

This is the code that I am currently working with

Expected behavior 🔧

I expected the plot to only eliminate the rank I clicked on not make the whole plot disappear.

Screenshots 📸

When I selected the rank marked in red the entire plot disappeared

Environment information: (please complete the following information)

• OS: Ubuntu 22.04.3
• Python environment:
  • Python version: 3.10.12
  • plotly-resampler environment: e.g.: Jupyter(lab), Dash web app (which browser): Visual Studio Code, Google Chrome (web browser)
• plotly-resampler version:0.10.0

Additional context
Add any other context about the problem here.

[jonasvdd]

@aparajit7, could you share a minimal csv file, similar to sample-enlarged.csv, for which this behavior occurs?

[aparajit7]

Hello Jonas, I have shared the sample-enlarged.csv here as a Google Drive link. Can you please let me know if you have received it? Thanks, Aparajit sample-enlarged.csv <https://drive.google.com/file/d/1-iUpQd5vDXZQLw9ucvAtuy8g2yTyAbKw/view?usp=drive_web>

…

On Thu, Sep 5, 2024 at 7:02 AM Jonas Van Der Donckt < ***@***.***> wrote: @aparajit7 <https://github.com/aparajit7>, could you share a minimal csv file, similar to sample-enlarged.csv, for which this behavior occurs? — Reply to this email directly, view it on GitHub <#317 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALVR5DM5AQFINWMRKZKNHW3ZVA23PAVCNFSM6AAAAABNTITVRCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMZRGIZTEMJSGQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jonasvdd]

@aparajit7 - The access was denied, so I sent an access request.

[aparajit7]

Hey Jonas, I have added the access. Can you check it now?

…

On Fri, Sep 6, 2024 at 5:20 AM Jonas Van Der Donckt < ***@***.***> wrote: @aparajit7 <https://github.com/aparajit7> - The access was denied, so I sent an access request. — Reply to this email directly, view it on GitHub <#317 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALVR5DNAUJ6HBXALRXFYUQDZVFXUFAVCNFSM6AAAAABNTITVRCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMZTGYZTCNBWHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jonasvdd]

@aparajit7,

It appears to be related to the error_x attribute. If you leave this out, everything works as expected.
Plotly-Resampler does not (yet) support this attribute.

[aparajit7]

Hello Jonas, I have tried to disable the error_x attribute but its actually not giving me the required output. Is there a way to use any other attribute or function similar to it so that Plotly-Resampler works without issues while displaying the ranks?

…

On Mon, Sep 9, 2024 at 3:14 AM Jonas Van Der Donckt < ***@***.***> wrote: @aparajit7 <https://github.com/aparajit7>, It appears to be related to the error_x attribute. If you leave this out, everything works as expected. Plotly-Resampler does not (yet) support this attribute. — Reply to this email directly, view it on GitHub <#317 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALVR5DOAKWLMQN7ZFOEAOWDZVVDF7AVCNFSM6AAAAABNTITVRCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMZXGMYTAMZXG4> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[aparajit7]

Hello Jonas, I don’t know if you received my previous email but I mentioned that disabling the error_x attribute would not work as it would change the plot completely. So is there any workaround to this? I would appreciate for the help. Thanks! On Sun, Sep 15, 2024 at 9:20 PM Aparajit Talukdar ***@***.***> wrote:

…

Hello Jonas, I have tried to disable the error_x attribute but its actually not giving me the required output. Is there a way to use any other attribute or function similar to it so that Plotly-Resampler works without issues while displaying the ranks? On Mon, Sep 9, 2024 at 3:14 AM Jonas Van Der Donckt < ***@***.***> wrote: > @aparajit7 <https://github.com/aparajit7>, > > It appears to be related to the error_x attribute. If you leave this > out, everything works as expected. > Plotly-Resampler does not (yet) support this attribute. > > — > Reply to this email directly, view it on GitHub > <#317 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ALVR5DOAKWLMQN7ZFOEAOWDZVVDF7AVCNFSM6AAAAABNTITVRCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMZXGMYTAMZXG4> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >