Example

Showcasing the awesome stuff made possible by boopiter

import matplotlib.pyplot as plt
import tqdm.auto as tqdm

Plots

Hey, could you generate some really brief simple code that would plot a 2D heat map with contour lines for some function you might make up? Matplotlib is fine.

🤖Reply🤖

Reply details

• Model: qwen2.5-coder:latest
• Finish reason: stop
• Tokens: 1204 prompt + 244 completion = 1448 total

Certainly! Here’s a snippet of Python code that generates a 2D heatmap with contour lines using Matplotlib:

import numpy as np
import matplotlib.pyplot as plt

# Define the function to plot
def f(x, y):
    return np.sin(np.sqrt(x**2 + y**2))

# Generate data
x = np.linspace(-5, 5, 100)
y = np.linspace(-5, 5, 100)
X, Y = np.meshgrid(x, y)
Z = f(X, Y)

# Create the plot
plt.figure(figsize=(8, 6))
plt.contourf(X, Y, Z, levels=20, cmap='viridis')
plt.colorbar(label='f(x, y)')
plt.contour(X, Y, Z, colors='black', levels=10)
plt.xlabel('x')
plt.ylabel('y')
plt.title('2D Heat Map with Contour Lines')
plt.show()

This code defines a simple function f and then uses Matplotlib to create a heatmap of this function along with contour lines. You can run this in your Python environment to see the plot.

import numpy as np
import matplotlib.pyplot as plt

# Define the function to plot
def f(x, y):
    return np.sin(np.sqrt(x**2 + y**2))

# Generate data
x = np.linspace(-5, 5, 100)
y = np.linspace(-5, 5, 100)
X, Y = np.meshgrid(x, y)
Z = f(X, Y)

# Create the plot
plt.figure(figsize=(8, 6))
plt.contourf(X, Y, Z, levels=20, cmap='viridis')
plt.colorbar(label='f(x, y)')
plt.contour(X, Y, Z, colors='black', levels=10)
plt.xlabel('x')
plt.ylabel('y')
plt.title('2D Heat Map with Contour Lines')
plt.show()

<ipython-input-1-6356375be1f3>:22: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown
  plt.show()

Progress Bars

We can only use the regular TQDM as if we were in a terminal because the notebook version requires IPy widgets that I’m not going to implement

from tqdm import tqdm
import time

for i in tqdm(range(1000)):
    time.sleep(0.01)  # Simulate some work being done

100%|##########| 1000/1000 [00:13<00:00, 72.47it/s]

Notice how the notebook version isn’t going to increment. So be aware.

from tqdm.notebook import tqdm
import time

for i in tqdm(range(1000)):
    time.sleep(0.01)  # Simulate some work being done

  0%|          | 0/1000 [00:00<?, ?it/s]

…See it just sits there and does nothing. So don’t try to use the notebook version.

Good news is, if you just use the auto version of TQDM, you get the regular text version:

from tqdm.auto import tqdm
import time

for i in tqdm(range(1000)):
    time.sleep(0.01)  # Simulate some work being done

100%|##########| 1000/1000 [00:12<00:00, 82.16it/s]

Which variety of which LLM are you?

🤖Reply🤖

Reply details

• Model: qwen2.5-coder:latest
• Finish reason: stop
• Tokens: 1866 prompt + 63 completion = 1929 total

I am Qwen, a large language model created by Alibaba Cloud. My base architecture is based on transformer models and I have been trained on a diverse dataset to understand and generate human-like text on a wide range of topics. If you have any questions or need assistance with anything specific, feel free to ask!

I’m wondering if you could give me a little code that would generate like a 3D scatter plot in Plotly that I can test here in this Boopiter notebook. I suppose I’ll have to render it via the IPython display HTML element, so give me your best guess at what that needs to look like. And please conform to my coding style of keeping everything as vertically compact as possible so that I don’t have to scroll up and down much to see the full code

🤖Reply🤖

Reply details

• Model: Claude Opus 4.8 (dead drop)

Compact as requested. One change with a purpose, though: I wrapped it in an <iframe srcdoc=...> instead of a bare display(HTML(fig.to_html(...))).

import numpy as np, plotly.graph_objs as go
from IPython.display import display, HTML
np.random.seed(42); n = 100
x, y = np.random.rand(2, n)*2 - 1; z = np.hypot(x, y)
fig = go.Figure(go.Scatter3d(x=x, y=y, z=z, mode='markers',
    marker=dict(size=5, color=z, colorscale='Viridis', opacity=.8)))
fig.update_layout(title='3D Scatter', margin=dict(l=0, r=0, b=0, t=30),
    scene=dict(xaxis_title='X', yaxis_title='Y', zaxis_title='Z'))
h = fig.to_html(include_plotlyjs='cdn', full_html=True).replace('&','&').replace('"','"')
display({'text/html': f'<iframe srcdoc="{h}" width="100%" height="500" style="border:none"></iframe>'}, raw=True)</iframe>'))

Why the iframe: raw Plotly HTML leads with <script> tags, and the docs generator (Quarto/Pandoc) lifts a raw-HTML block that starts with <script> up out of the page flow – which is why, on the doc site, the plot detaches from its cell and jumps to the top (and two plots would collide up there). An <iframe> is a single inline element with the whole plot sealed inside its srcdoc, so there’s nothing to hoist – it renders right under the code, in the notebook and on the docs page.

Bonus: include_plotlyjs='cdn' pulls plotly.js from the CDN instead of inlining ~4.8MB of library into every output – so the cell (and the committed notebook) stays tiny.

import numpy as np, plotly.graph_objs as go
from IPython.display import display
np.random.seed(42); n = 100
x, y = np.random.rand(2, n)*2 - 1; z = np.hypot(x, y)
fig = go.Figure(go.Scatter3d(x=x, y=y, z=z, mode='markers',
    marker=dict(size=5, color=z, colorscale='Viridis', opacity=.8)))
fig.update_layout(title='3D Scatter', margin=dict(l=0, r=0, b=0, t=30),
    scene=dict(xaxis_title='X', yaxis_title='Y', zaxis_title='Z'))
h = fig.to_html(include_plotlyjs='cdn', full_html=True).replace('&','&').replace('"','"')
display({'text/html': f'<iframe srcdoc="{h}" width="100%" height="500" style="border:none"></iframe>'}, raw=True)

^^Go ahead! Manipulate that plot with your mouse. Press & drag, zoom in, etc.