---
jupytext:
  text_representation:
    extension: .md
    format_name: myst
    format_version: 0.13
    jupytext_version: 1.17.0
---

+++ {"tags": ["jupyterlite_sphinx_strip"]}

<!-- Here, we specify the NotebookLite directive from jupyterlite-sphinx:
https://jupyterlite-sphinx.readthedocs.io/en/stable/directives/notebooklite.html

This directive is used to include a JupyterLite notebook in the documentation using
the Notebook interface. We'll use the "new tab" option to create a button that will
open the JupyterLite deployment with it, and customise the button text.

The button does not contain a style by default. You may use CSS to style it as per
your needs and requirements, according to the theme used. We describe it here:
https://jupyterlite-sphinx.readthedocs.io/en/stable/directives/try_examples.html#configuration

For the purposes of this demo, we have provided a custom CSS file in the
`docs/source/_static` directory, which is included in the Sphinx build process.

If the strip_tagged_cells configuration option is set in conf.py, any cell that is
wrapped in the `jupyterlite_sphinx_strip` tag will be stripped from the final output,
so that it won't be included in the JupyterLite deployment.
-->

```{eval-rst}
.. notebooklite:: matplotlib_demo.md
    :new_tab: True
    :new_tab_button_text: Try it online!
```

+++

# Matplotlib interactive demo

## Plotting in JupyterLite

The packages used in this notebook are NumPy and `matplotlib`, which are available in [the emscripten-forge channel](https://beta.mamba.pm/channels/emscripten-forge).
They are pre-installed into the deployment using the `ennvironment.yml` file in the docs source directory for the Xeus kernel to load. Once the kernel is started, the packages are available for use right away.

Any extra packages must be pre-installed into the environment before the JupyterLite deployment is built as a part of the Sphinx build process.

```{code-cell}
import matplotlib.pyplot as plt
import numpy as np

np.random.seed(0)

n = 100

x = list(range(n))
y = np.cumsum(np.random.randn(n)) + 100.

plt.figure(figsize=(10, 6))
plt.plot(x, y)
plt.xlabel('Index')
plt.ylabel('Value')
plt.title('Lines')
plt.grid(True)
plt.show()
```

```{code-cell}
# Update the line plot with green color, markers and fill
plt.figure(figsize=(10, 6))
plt.plot(x, y, color='green', marker='o')
plt.fill_between(x, y, min(y), color='green', alpha=0.3)
plt.xlabel('Index')
plt.ylabel('Value')
plt.title('Lines')
plt.grid(True)
plt.show()
```

```{code-cell}
n = 100

x = list(range(n))
y = np.cumsum(np.random.randn(n))

plt.figure(figsize=(10, 6))
plt.bar(x, y)
plt.xlabel('Index')
plt.ylabel('Value')
plt.title('Bars')
plt.grid(True, axis='y')
plt.show()
```

```{code-cell}
# Update the bars with new random data
new_y = np.cumsum(np.random.randn(n))
plt.figure(figsize=(10, 6))
plt.bar(x, new_y)
plt.xlabel('Index')
plt.ylabel('Value')
plt.title('Bars')
plt.grid(True, axis='y')
plt.show()
```

### Multiple line plots

```{code-cell}
import numpy as np
import matplotlib.pyplot as plt

np.random.seed(0)
y1 = np.cumsum(np.random.randn(150)) + 100.
y2 = np.cumsum(np.random.randn(150)) + 100.
y3 = np.cumsum(np.random.randn(150)) + 100.
y4 = np.cumsum(np.random.randn(150)) + 100.

x = np.arange(len(y1))

plt.figure(figsize=(12, 7))
plt.plot(x, y1, label='Series 1')
plt.plot(x, y2, label='Series 2')
plt.plot(x, y3, label='Series 3')
plt.plot(x, y4, label='Series 4')
plt.xlabel('Index')
plt.ylabel('Value')
plt.title('Lines')
plt.legend()
plt.grid(True)
plt.show()
```