-
-
Notifications
You must be signed in to change notification settings - Fork 2.7k
Open
Description
Steps to reproduce:
- create an environemnt with plotly==6.2.0 + numpy
- run this code (purely random plot), note the
write_json
at the end
import numpy as np
# Generate random data
np.random.seed(42) # for reproducibility
random_x = np.random.rand(100) * 10 # 100 random numbers between 0 and 10
random_y = np.random.rand(100) * 10 # 100 random numbers between 0 and 10
# Create a scatter trace
scatter_trace = go.Scatter(
x=random_x,
y=random_y,
mode='markers', # Display markers for each data point
marker=dict(size=8, color='blue', opacity=0.7) # Customize marker appearance
)
# Create a Figure object and add the trace
fig = go.Figure(data=[scatter_trace])
# Update layout (optional)
fig.update_layout(
title='Random Scatter Plot (go.Figure)',
xaxis_title='X-axis',
yaxis_title='Y-axis'
)
# Display the plot
fig.show()
fig.write_json("test1.json")
- create a second environment with plotly<6.0.0 (5.9.0 in my example).
- run this code
from plotly.io import read_json
content = read_json("test1.json")
# content = read_json("test1.json", skip_invalid="True")
content.show()
- observe the next error:
Invalid value of type 'builtins.dict' received for the 'x' property of scatter
Received value: {'dtype': 'f8', 'bdata': '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'}
The 'x' property is an array that may be specified as a tuple,
list, numpy array, or pandas Series
Basically, figures saved with Plotly 6.x are not read back correctly when using Plotly 5.x. I am not if this is how it's supposed to be or not.
If I try content = read_json("test1.json", skip_invalid="True")
it doesn't produce an error but the plot is completely wrong.
I would love any feedback on this, thanks in advance.
Metadata
Metadata
Assignees
Labels
No labels