Skip to content

README.md

Latest commit

 

History

History
184 lines (137 loc) · 5.21 KB

README.md

File metadata and controls

184 lines (137 loc) · 5.21 KB

Overview

Kaleido is a cross-platform Python library for generating static images (e.g. png, svg, pdf, etc.) for Plotly.js, to be used by Plotly.py.

Installation

Kaleido can be installed from PyPI using pip:

$ pip install kaleido --upgrade

As of version 1.0.0, Kaleido requires Chrome to be installed. If you already have Chrome on your system, Kaleido should find it; otherwise, you can install a compatible Chrome version using the kaleido_get_chrome command:

$ kaleido_get_chrome

or function in Python:

import kaleido
kaleido.get_chrome_sync()

Migrating from v0 to v1

Kaleido v1 introduces a new API. If you're currently using v0, you'll need to make changes to your code and environment where you are running Kaleido.

  • If using Kaleido v1 with Plotly.py, you will need to install Plotly.py v6.1.1 or later.
  • Chrome is no longer included with Kaleido. Kaleido will look for an existing Chrome installation, but also provides commands for installing Chrome. If you don't have Chrome, you'll need to install it. See the installation section above for instructions.
  • If your code uses Kaleido directly: kaleido.scopes.plotly has been removed in v1. Kaleido v1 provides write_fig and write_fig_sync for exporting Plotly figures.
from kaleido import write_fig_sync
import plotly.graph_objects as go

fig = go.Figure(data=[go.Scatter(y=[1, 3, 2])])
kaleido.write_fig_sync(fig, path="figure.png")

Development guide

Below are examples of how to use Kaleido directly in your Python program.

If you want to export images of Plotly charts, it's not necessary to call Kaleido directly; you can use functions in the Plotly library. See the Plotly documentation for instructions.

Usage examples

import kaleido

async with kaleido.Kaleido(n=4, timeout=90) as k:
  # n is number of processes
  await k.write_fig(fig, path="./", opts={"format":"jpg"})

# other `kaleido.Kaleido` arguments:
# page:  Change library version (see PageGenerators below)

# `Kaleido.write_fig()` arguments:
# - fig:       A single plotly figure or an iterable.
# - path:      A directory (names auto-generated based on title)
#              or a single file.
# - opts:      A dictionary with image options:
#              `{"scale":..., "format":..., "width":..., "height":...}`
# - error_log: If you pass a list here, image-generation errors will be appended
#              to the list and generation continues. If left as `None`, the
#              first error will cause failure.

# You can also use Kaleido.write_fig_from_object:
  await k.write_fig_from_object(fig_objects, error_log)
# where `fig_objects` is a dict to be expanded to the fig, path, opts arguments.

There are shortcut functions which can be used to generate images without creating a Kaleido() object:

import asyncio
import kaleido
asyncio.run(
  kaleido.write_fig(
    fig,
    path="./",
    n=4
  )
)

You can also use threaded and synchronous Python. Here is an example how to create one Kaleido instance per-application, and launch the background Chrome browser for rendering only once.

import threading
from io import BytesIO
import asyncio

from plotly.graph_objects import Figure
from kaleido import Kaleido

_kaleido = None
_lock = threading.Lock()


def get_kaleido() -> Kaleido:
  """Create Kaleido rendering backend.

  - Creates a Chrome browser on background

  - `See usage examples <https://github.com/plotly/Kaleido>`__
  """
  global _kaleido
  with _lock:
    if _kaleido is None:
      _kaleido = Kaleido()

  return _kaleido


def render_plotly_figure_as_image_file(
  figure: Figure,
  format: str = "png",
  width: int = 512,
  height: int = 512,
) -> bytes:
  """"Render Plotly figure as a static PNG image.

  - Uses Kaleido to render the Plotly figure as a PNG, SVG image, PDF or similar.
  - Creates Kaleido backend (Chrome browser) on the first call.

  :param format:
    See ``kaleido._fig_tools`` module for supported forms.

  :param width:
    Width in pixels

  :param height:
    Height in pixels

  :return:
    Image data encoded as bytes blob.
  """

  stream = BytesIO()

  kaleido_instance = get_kaleido()

  opts = dict(
    format=format,
    width=width,
    height=height,
  )

  asyncio.run(
    kaleido_instance.write_fig(
      figure,
      stream,
      opts=opts,
    )
  )
  data = stream.getvalue()
  assert len(data) > 0, "Rendered image data is empty"
  stream.close()
  return data

PageGenerators

The page argument takes a kaleido.PageGenerator() to customize versions. Normally, kaleido looks for an installed plotly as uses that version. You can pass kaleido.PageGenerator(force_cdn=True) to force use of a CDN version of plotly (the default if plotly is not installed).

my_page = kaleido.PageGenerator(
  plotly="A fully qualified link to plotly (https:// or file://)",
  mathjax=False # no mathjax, or another fully quality link
  others=["a list of other script links to include"]
)