You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Binary transport will potentially lower the size of the arrays being passed to JS, and might speed up the process as well.
How we could implement it: Given a numpy array, we compute the shape, the dtype, the b64 encoded memory view of the flattened numpy array. Then those three values are passed as a dict/json to the frontend, which is then reconstructed into a typedarray. The plan would be to use this function on the python side.
Binary transport will potentially lower the size of the arrays being passed to JS, and might speed up the process as well.
How we could implement it: Given a numpy array, we compute the shape, the dtype, the b64 encoded memory view of the flattened numpy array. Then those three values are passed as a dict/json to the frontend, which is then reconstructed into a typedarray. The plan would be to use this function on the python side.
For more information, see this pull request as well as plotly/plotly.py#2943
The text was updated successfully, but these errors were encountered: