research.notebook

Documentation for eth_defi.research.notebook Python module.

Jupyter notebook set up and formatting utilities.

Functions

display_dataframe_with_html(df)	Display a DataFrame in a notebook with clickable HTML links.
format_large_number(n)	Format a large number with k/M/B suffixes.
set_large_plotly_chart_font([...])	Increase the default Plotly chart font sizes so that charts are readable on other mediums like mobile and PowerPoint.
set_notebook_logging([log_level])	Enable logging in notebooks.
setup_charting_and_output([mode, ...])	Sets charting and other output options for Jupyter Notebooks.

Classes

OutputMode

What is the output mode for the notebook visualisations.

class OutputMode

Bases: enum.Enum

What is the output mode for the notebook visualisations.

Interactive visualisations work only on the HTML pages that are able to load Plotly.js JavaScripts.

For examples see setup_charting_and_output().

display_dataframe_with_html(df)

Display a DataFrame in a notebook with clickable HTML links.

Use instead of display(df) when the DataFrame contains HTML content such as <a> links that should be rendered.

Example:

from eth_defi.research.notebook import display_dataframe_with_html

display_dataframe_with_html(stats)

Parameters

df (pandas.DataFrame) –

format_large_number(n)

Format a large number with k/M/B suffixes.

Example:

>>> format_large_number(1_234_567)
'1.2M'
>>> format_large_number(45_678)
'45.7k'
>>> format_large_number(999)
'999'

Parameters

n (Optional[Union[float, int]]) –

Return type

str

set_large_plotly_chart_font(title_font_size=30, font_size=24, legend_font_size=24, line_width=3, axis_title_font_size=24, base_template='plotly')

Increase the default Plotly chart font sizes so that charts are readable on other mediums like mobile and PowerPoint.

Usage:

from tradeexecutor.utils.notebook import set_large_plotly_chart_font

set_large_plotly_chart_font()

set_notebook_logging(log_level=20)

Enable logging in notebooks.

• Only needed to diagnose Client library bugs when running in notebook

Parameters

log_level (Union[int, str]) –

setup_charting_and_output(mode=OutputMode.interactive, image_format='svg', max_rows=1000, width=1500, height=1500, increase_font_size=False)

Sets charting and other output options for Jupyter Notebooks.

Interactive charts are better for local development, but are not compatible with most web-based notebook viewers.

• Set Quantstats chart to SVG output and for high-resolution screens

• Mute common warnings like Matplotlib font loading

• Plotly discussion

Example how to set up default interactive output settings. Add early of your notebook do:

# Set Jupyter Notebook output mode parameters.
# For example, table max output rows is lifted from 20 to unlimited.
from tradeexecutor.utils.notebook import setup_charting_and_output

setup_charting_and_output()

Example how to set up static image rendering:

# Set charts to static image output, 1500 x 1000 pixels from tradeexecutor.utils.notebook import setup_charting_and_output, OutputMode setup_charting_and_output(OutputMode.static, image_format=”png”, width=1500, height=1000)

Parameters

• mode (eth_defi.research.notebook.OutputMode) – What kind of viewing context we have for this notebook output

• image_format –

  Do we do SVG or PNG.

  SVG is better, but Github inline viewer cannot display it in the notebooks.

• max_rows –

  Do we remove the max_rows limitation from Pandas tables.

  Default 20 is too low to display summary tables.

• increase_font_size – Make charts and tables more readable with larger fonts