-
-
Notifications
You must be signed in to change notification settings - Fork 2.7k
Open
Labels
P1needed for current cycleneeded for current cyclebugsomething brokensomething brokenregressionthis used to workthis used to work
Description
the x-axis of my graph is displaying as the ns value of the timestamp, 1715000000000000000, instead of the corresponding readable timestamp.
The code is
fig = go.Figure()
local_df["reservation_start_week_utc"] = pd.to_datetime(local_df[
"reservation_start_week_utc"
].apply(lambda x: pd.Timestamp(x).to_pydatetime()))
fig.add_trace(
go.Scatter(
x=local_df["reservation_start_week_utc"],
y=local_df["weighted_average_price"],
...
the datatype of the column that populates my x-axis is datetime64[ns, UTC]
I tried casting to different types, and also several different format settings on the plot. I couldnt get it to work. Rolling back plotly immediately fixed the bug
python versions used 3.13 and 3.10 no difference.
rolling back from plotly 6.0.1 to 5.24.1 fixed the formatting issue
TermeHansen and miohtama
Metadata
Metadata
Assignees
Labels
P1needed for current cycleneeded for current cyclebugsomething brokensomething brokenregressionthis used to workthis used to work