Values not plotted in right order for large dataframes

[rik-vandervlist-sympower]

Environment

• Operating System: Linux (databricks runtime 11.3 ML)
• Python Version: 3.9.5
• How did you install bamboolib: pip
• Python packages: See Databricks runtime link

Description of Issue

• What did you expect to happen?
  To see a timeseries line graph of my data
• What happened instead?
  Lines of the graph go everywhere, since the plotted data is not properly sorted by timestamp

Reproduction Steps

1. Find a large dataset
2. Make a plot with timestamp column vs. a variable of interest
3. Graph shows up mangled.

The generated code looks as follows:

import plotly.express as px
fig = px.line(power_data.sort_values(by=['utc_timestamp'], ascending=[True]).sample(n=10000, replace=False, random_state=123).sort_index(), x='utc_timestamp', y='value', color='id')
fig

Note here that .sort_values(by=['utc_timestamp'], ascending=[True]) is placed before .sample(n=10000, replace=False, random_state=123). There is a .sort_index(), but the time column is not the index of my pandas dataframe. I think placing the sort_values after the random sample might resolve this issue

What steps have you taken to resolve this already?

See the above!

Anything else?