Chartsworth

Chartsworth posts from notebooks to slack. This little metric monster lets you post plots, images, and anything else the slack sdk will let you post.

Read more background and workflows around scheduling in the Notebook Reporting with Chart Monsters Post.

Installation and Setup

pip install chartsworth

Set environment variable CHARTSWORTH_TOKEN to your Slack App. You can create a new Slack App Create a new Slack App and learn more about creating a new slack app in the chartsworth post.

Usage

from chartsworth import Chartsworth
chartsworth = Chartsworth("#chartsy")

new_users = 11_003
chartsworth.post(f"We have {new_users} new users")

Note: many times you'll need to use a channel ID (e.g. C05HP8Z5ZPD), especially if the channel is private. Otherwise, images will not post (only text will).

Posting Plots

Chartsworth can take any PIL Image or Matplotlib Figure and post it to slack. This means you can post plots, images, and anything else you can think of.

from chartsworth import Chartsworth

chartsworth = Chartsworth("C05HP8Z5ZPD")
chartsworth.post("Who's ready for a 🏞️ stream plot? 🧵")

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(-3, 3, 100)
y = np.linspace(-3, 3, 100)
X, Y = np.meshgrid(x, y)
u = -1 - X**2 + Y
v = 1 + X - Y**2
speed = np.sqrt(u*u + v*v)

## Create a figure we can pass to Chartsworth ##
fig, ax = plt.subplots()

strm = ax.streamplot(X, Y, u, v, color=speed, linewidth=2, cmap='autumn')
fig.colorbar(strm.lines)

## Post!
chartsworth.post(fig)

Scheduling

Schedule your notebooks on Noteable from the notebooks UI

Or run it with papermill with your own custom scheduling!