Execute and cache your pages

Jupyter Book can automatically run and cache any notebook pages. Notebooks can either be run each time the documentation is built, or cached locally so that notebooks will only be re-run when the code cells in a notebook have changed.

Caching behaviour is controlled with the execute: section in your _config.yml file. See the sections below for each configuration option and its effect.

Tip

If you’d like to execute code that is in your Markdown files, you can use the {code-cell} directive in MyST Markdown. See Notebooks written entirely in Markdown for more information.

Trigger notebook execution

By default, Jupyter Book will execute any content files that have a notebook structure and that are missing at least one output. This is equivalent to the following configuration in _config.yml`:

execute:
  execute_notebooks: auto

This will only execute notebooks that are missing at least one output. If the notebook has all of its outputs populated, then it will not be executed.

To force the execution of all notebooks, regardless of their outputs, change the above configuration value to:

execute_notebooks: force

To cache execution outputs with jupyter-cache, change the above configuration value to:

execute:
  execute_notebooks: cache

See Caching the notebook execution for more information.

To turn off notebook execution, change the above configuration value to:

execute:
  execute_notebooks: 'off'

Exclude files from execution

To exclude certain file patterns from execution, use the following configuration:

execute:
  exclude_patterns:
    - 'pattern1'
    - 'pattern2'
    - '*pattern3withwildcard'

Any file that matches one of the items in exclude_patterns will not be executed.

Caching the notebook execution

You may also cache the results of executing a notebook page using jupyter-cache. In this case, when a page is executed, its outputs will be stored in a local database. This allows you to be sure that the outputs in your documentation are up-to-date, while saving time avoiding unnecessary re-execution. It also allows you to store your .ipynb files in your git repository without their outputs, but still leverage a cache to save time when building your site.

When you re-build your site, the following will happen:

  • Notebooks that have not seen changes to their code cells since the last build will not be re-executed. Instead, their outputs will be pulled from the cache and inserted into your site.

  • Notebooks that have had any change to their code cells will be re-executed and the cache will be updated with the new outputs.

To enable caching of notebook outputs, use the following configuration:

execute:
  execute_notebooks: cache

By default, the cache will be placed in the parent of your build folder. Generally, this is in _build/.jupyter_cache.

You may also specify a path to the location of a jupyter cache you’d like to use:

execute:
  cache: path/to/mycache

The path should point to an empty folder, or a folder where a jupyter cache already exists.

Execution configuration

You can control notebook execution and how output content is handled at a project level using your _config.yml but, in some cases, also at a notebook and code cell level. Below we explore a number of ways to achieve this.

The execution working directory

Important

The default behaviour of cache is now to run in the local directory. This is a change from v0.7.

By default, the command working directory (cwd) in which a notebook runs will be the directory in which it is located (for both auto and cache). This means that notebooks requiring access to assets in relative paths will work.

Alternatively, if you wish for your notebooks to isolate your notebook execution in a temporary folder, you can use the following _config.yml setting:

execute:
  run_in_temp: true

Setting execution timeout

Execution timeout defines the maximum time (in seconds) each notebook cell is allowed to run for. If the execution takes longer an exception will be raised. The default is 30 seconds, so in cases of long-running cells you may want to specify a higher value. The timeout option can also be set to -1, to remove any restriction on execution time.

You can set the timeout for all notebook executions in your _config.yml:

execute:
  timeout: 100

This global value can also be overridden per notebook by adding this to your notebook metadata:

{
 "metadata": {
  "execution": {
      "timeout": 30
  }
}

Dealing with code that raises errors

In some cases, you may want to intentionally show code that doesn’t work (e.g., to show the error message).

You can allow errors for all notebooks in your _config.yml:

execute:
  allow_errors: true

This global value can also be overridden per notebook by adding this to your notebook metadata:

{
 "metadata": {
  "execution": {
      "allow_errors": false
  }
}

Lastly, you can allow errors at a cell level, by adding a raises-exception tag to your code cell. This can be done via a Jupyter interface, or via the {code-cell} directive like so:

```{code-cell}
---
tags: [raises-exception]
---
print(thisvariabledoesntexist)
```

Which produces:

print(thisvariabledoesntexist)
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-1-125c53ec1b82> in <module>
----> 1 print(thisvariabledoesntexist)

NameError: name 'thisvariabledoesntexist' is not defined

Dealing with code that produces stderr

You may also wish to control how stderr outputs are dealt with.

Alternatively, you can configure how stdout is dealt with at a global configuration level, using the nb_output_stderr configuration value.

You can configure the default behaviour for all notebooks in your _config.yml:

execute:
  stderr_output: show

Where the value is one of:

  • "show" (default): show all stderr (unless a remove-stderr tag is present)

  • "remove": remove all stderr

  • "remove-warn": remove all stderr, but log a warning if any found

  • "warn", "error" or "severe": log the stderr at a certain level, if any found.

You can also remove stderr at a cell level, using the remove-stderr cell tag, like so:

```{code-cell} ipython3
:tags: [remove-stderr]

import sys
print("this is some stdout")
print("this is some stderr", file=sys.stderr)
```

which produces

import sys
print("this is some stdout")
print("this is some stderr", file=sys.stderr)
this is some stdout

Dealing with code that produces stdout

Similar to stderr, you can remove stdout at a cell level with the remove-stdout tag, by which

```{code-cell} ipython3
:tags: [remove-stdout]

import sys
print("this is some stdout")
print("this is some stderr", file=sys.stderr)
```

produces the following:

import sys
print("this is some stdout")
print("this is some stderr", file=sys.stderr)
this is some stderr

Execution statistics

As notebooks are executed, certain statistics are stored on the build environment by MyST-NB. The simplest way to access and visualise this data is using the {nb-exec-table} directive.

See also

The MyST-NB documentation, for creating your own directives to manipulate this data.

The simple directive

```{nb-exec-table}
```

produces:

Document

Modified

Method

Run Time (s)

Status

content/citations

2020-10-26 15:25

cache

2.02

content/code-outputs

2020-10-26 15:25

cache

5.03

content/execute

2020-10-26 15:25

cache

1.17

content/layout

2020-10-26 15:25

cache

2.0

content/math

2020-10-26 15:25

cache

2.02

file-types/jupytext

2020-10-26 15:25

cache

0.96

file-types/myst-notebooks

2020-10-26 15:25

cache

2.02

file-types/notebooks

2020-10-26 15:25

cache

3.8

interactive/hiding

2020-10-26 15:25

cache

3.02

interactive/interactive

2020-10-26 15:25

cache

2.47

interactive/launchbuttons

2020-10-26 15:25

cache

1.61

reference/cheatsheet

2020-10-26 15:25

cache

3.9

start/overview

2020-10-26 15:26

cache

2.81