Sandbox - Interactive Python Execution

Test the interactive Python execution with matplotlib, pandas, and Bokeh.

Python built-in libraries

Basic example

# Testing built-in library import
import math
import datetime

# Testing printing
print("Hello from DataViz!")
print("Current date and time:", datetime.datetime.now())

# Testing basic logic
primes_below_100 = [
  x 
  for x in range(2, 100) 
  if all(x % y != 0 for y in range(2, int(math.sqrt(x)) + 1))
]
print("All the prime numbers below 100 are:")
print(primes_below_100)

🔴 Inputs Not working

Nagini timeout should stop execution after 30 seconds

try:
  x = input()
  print(f"You entered: {x}")
except Exception as e:
  print(f"Error: {e}")

`matplotlib` (+ `pandas`) examples

Simple `matplotlib` example

import matplotlib.pyplot as plt
import numpy as np
import warnings
import matplotlib

# Suppress matplotlib font warnings for cleaner output
warnings.filterwarnings('ignore', message='findfont: Generic family')
matplotlib.rcParams['font.family'] = 'DejaVu Sans'  # Use available fallback font

# Create data
x = np.linspace(0, 2 * np.pi, 100)
y_sin = np.sin(x)
y_cos = np.cos(x)

# Create the plot
plt.figure(figsize=(10, 6))
plt.plot(x, y_sin, 'b-', label='sin(x)', linewidth=2)
plt.plot(x, y_cos, 'r--', label='cos(x)', linewidth=2)
plt.xlabel('x')
plt.ylabel('y')
plt.title('Sine and Cosine Functions')
plt.legend()
plt.grid(True, alpha=0.3)
plt.show()

print("Plot generated successfully!")

`pandas`-only example

import pandas as pd
import numpy as np

# Create a sample DataFrame
data = {
    'Name': ['Alice', 'Bob', 'Charlie', 'David', 'Eve'],
    'Age': [25, 30, 35, 28, 32],
    'Score': [85, 92, 78, 91, 88],
    'Department': ['IT', 'HR', 'IT', 'Finance', 'HR']
}

df = pd.DataFrame(data)

print("Original DataFrame:")
print(df)
print("\n" + "="*50 + "\n")

# Basic statistics
print("Statistics:")
print(df.describe())
print("\n" + "="*50 + "\n")

# Group by department
print("Average score by department:")
dept_avg = df.groupby('Department')['Score'].mean()
print(dept_avg)

`pandas` + `matplotlib` time series example

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import warnings
import matplotlib

# Suppress matplotlib font warnings for cleaner output
warnings.filterwarnings('ignore', message='findfont: Generic family')
matplotlib.rcParams['font.family'] = 'DejaVu Sans'  # Use available fallback font

# Create sample time series data
dates = pd.date_range('20240101', periods=100)
df = pd.DataFrame({
    'date': dates,
    'value': np.random.randn(100).cumsum() + 100,
    'volume': np.random.randint(50, 150, 100)
})

# Create subplots
fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 8))

# Plot time series
ax1.plot(df['date'], df['value'], 'b-', linewidth=2)
ax1.set_title('Time Series Data')
ax1.set_xlabel('Date')
ax1.set_ylabel('Value')
ax1.grid(True, alpha=0.3)

# Plot volume bars
ax2.bar(df['date'], df['volume'], color='green', alpha=0.6)
ax2.set_title('Daily Volume')
ax2.set_xlabel('Date')
ax2.set_ylabel('Volume')
ax2.grid(True, alpha=0.3)

plt.tight_layout()
plt.show()

print(f"Data shape: {df.shape}")
print(f"Date range: {df['date'].min()} to {df['date'].max()}")
print(f"Value range: {df['value'].min():.2f} to {df['value'].max():.2f}")

`bokeh` interactive plots

🟢 Working example (for `pm-codex`)

This code creates a fully interactive sine wave plot using Bokeh with hover tooltips, zoom/pan functionality, and clickable legends.

# 🚀 ORIGINAL VERSION - Works with your Pyodide wrapper
# 📦 Import Bokeh plotting + document
from bokeh.plotting import figure, curdoc  
# 🎯 Import hover functionality
from bokeh.models import HoverTool     
# 🔢 Import numerical operations     
import numpy as np                          

# 🔢 100 points from 0 to 4π (2 sine cycles)
x = np.linspace(0, 4*np.pi, 100)           
# 📈 Calculate sine values for each x
y = np.sin(x)                              

# 📝 Plot title: `title` argument
# 📏 Plot width in pixels: `width` argument
# 📏 Plot height in pixels: `height` argument
# 🛠️ Interactive tools available: `tools` argument
p = figure(
    title="Interactive Sine Wave - Try zooming and panning!",  
    width=600,                             
    height=400,                           
    tools="pan,wheel_zoom,box_zoom,reset"  
)

# 📊 Blue sine curve
p.line(x, y, legend_label="sin(x)", line_width=2, color='navy')  
# 🔴 Red dots every 5th point
p.scatter(x[::5], y[::5], size=8, color='red', alpha=0.5)      

# 💬 Add hover functionality: show coordinates on hover
hover = HoverTool(tooltips=[("(x,y)", "($x, $y)")])  
# ➕ Add hover tool to plot
p.add_tools(hover)                         

# 🎨 Customize plot appearance
# 📊 X-axis label
p.xaxis.axis_label = "x"                   
# 📊 Y-axis label
p.yaxis.axis_label = "sin(x)"             
# 👁️ Click legend to hide/show lines
p.legend.click_policy = "hide"             

# 📺 Display plot in document (`pm-codex`): `curdoc().add_root(p)`
curdoc().add_root(p)                       

# 📢 Success messages
print("✅ Interactive Bokeh plot created!")
print("🎯 Use mouse to pan, scroll to zoom, hover for values!")

🔴 Working example (for `jupyter`)

This code is designed to work in standard and local JupyterLab. It creates a fully interactive sine wave plot using Bokeh with hover tooltips, zoom/pan functionality, and clickable legends.

# 📚 JUPYTER VERSION - Works in standard JupyterLab
# 📦 Import plotting + show function
from bokeh.plotting import figure, show   
# ❌ OLD: curdoc not needed for Jupyter
# from bokeh.plotting import figure, curdoc 
# ✨ NEW: Enable notebook output mode
from bokeh.io import output_notebook    
# 🎯 Import hover functionality
from bokeh.models import HoverTool   

# 🔢 Import numerical operations
import numpy as np                          

# 🔧 CRITICAL: Specific NOTEBOOK: Enable Jupyter notebook display mode
# ✨ Specific NOTEBOOK: Tell Bokeh to render in notebook cells
output_notebook()                           

# 🔢 100 points from 0 to 4π (2 sine cycles)
x = np.linspace(0, 4*np.pi, 100)           
# 📈 Calculate sine values for each x
y = np.sin(x)                              

# 📝 Plot title
# 📏 Plot width in pixels
# 📏 Plot height in pixels
# 🛠️ Interactive tools available
p = figure(
    title="Interactive Sine Wave - Try zooming and panning!",  
    width=600,                             
    height=400,                            
    tools="pan,wheel_zoom,box_zoom,reset"  
)

# 📊 Blue sine curve
p.line(x, y, legend_label="sin(x)", line_width=2, color='navy')  
# 🔴 Red dots every 5th point
p.scatter(x[::5], y[::5], size=8, color='red', alpha=0.5)       

# 💬 Add hover functionality: show coordinates on hover
hover = HoverTool(tooltips=[("(x,y)", "($x, $y)")])  
# ➕ Add hover tool to plot
p.add_tools(hover)                         

# 🎨 Customize plot appearance
# 📊 X-axis label
p.xaxis.axis_label = "x"        
# 📊 Y-axis label
p.yaxis.axis_label = "sin(x)"       
# 👁️ Click legend to hide/show lines
p.legend.click_policy = "hide"             

# 📺 Display plot (Jupyter method)
# ✨ NEW: Specific NOTEBOOK: Display plot directly in cell output
show(p)        
# ❌ OLD: Not needed for Jupyter
# curdoc().add_root(p)                     

# 📢 Success messages
print("✅ Interactive Bokeh plot created!")
print("🎯 Use mouse to pan, scroll to zoom, hover for values!")

🟡 Deprecation warning (for `pm-codex`)

This example intentionally uses the deprecated circle() method to demonstrate the compact warning display.

from bokeh.plotting import figure, curdoc
from bokeh.models import HoverTool
import numpy as np

# Create simple data
x = [1, 2, 3, 4, 5]
y = [2, 5, 3, 8, 7]

# Create plot
p = figure(
    title="Deprecation Warning Demo",
    width=400, 
    height=300,
    tools="pan,wheel_zoom,reset"
)

# This will trigger the deprecation warning
p.circle(x, y, size=15, color='red', alpha=0.6)

# Add to document
curdoc().add_root(p)

print("⚠️  This example uses deprecated circle() method")
print("📏 Check the compact warning display below!")

`pm-codex` integration tests

With `height_in_px`

# CodeMirror will set the height to 250px for the Codex container
# Your Python code here
print("Hello World")

With `script_path`

# _________________foreground_script_________________


# _________________corrector_________________
from io import StringIO
from pprint import pprint
import sys
import typing
import unittest
class PythonCorrector:
    def __init__(self, background_script):
        self.background_script = background_script
        self.last_error: str | None = None
        self.testcase = unittest.TestCase()
        self.check_counter = 0
        self.correction = {}
        # TODO: move this in the teacher class
        # Create a StringIO object to capture output
        self.original_stdout = sys.stdout
        # self.stdout = StringIO()
        # sys.stdout = self.stdout
        self.stdout = sys.stdout

        # if pyodide
        # todo : write surcouche pour data get

        print(sys.platform)

        #     print(codex_id)
        #     # load the data-background-script with js
        #     self.background_script = ''
        #     # js.globals.data_background_script

        #     self.background_script = document.activeElement.id

    def __del__(self):
        sys.stdout = sys.__stdout__

    @staticmethod
    def gracefully_check(check_method):
        def wrapper(self, *args, **kwargs):
            try:
                check_method(self, *args, **kwargs)
                success, msg = 1, ''
            except AssertionError as e:
                success, msg = 0, str(e)

            self.check_counter += 1
            self.correction[self.check_counter] = (success, msg)
            return success, msg

        return wrapper

    def check_coincide(self, x: typing.Any) -> None:
        namespace = {}
        exec(self.background_script, namespace)
        # TODO : SAME SETUP FOR ALL BECAUSE OF AUTOFORMAT
        x_name = f'{x=}'.split('=')[0]
        self.check_equal(x, namespace[x_name])

    @gracefully_check
    def check_equal(self, x: typing.Any, y: typing.Any) -> None:
        # First check if types match
        self.testcase.assertIsInstance(x, type(y))
        # Booleans
        if isinstance(x, bool):
            self.testcase.assertEqual(x, y)
        # Strings
        if isinstance(x, str):
            self.testcase.assertEqual(x, y)
        # Numbers
        elif isinstance(x, int):
            self.testcase.assertEqual(x, y)
        elif isinstance(x, float):
            self.testcase.assertAlmostEqual(x, y)
        # Sequences
        elif isinstance(x, list):
            self.testcase.assertListEqual(x, y)
        elif isinstance(x, tuple):
            self.testcase.assertTupleEqual(x, y)
        elif isinstance(x, set):
            self.testcase.assertSetEqual(x, y)
        # Dictionaries
        elif isinstance(x, dict):
            self.testcase.assertDictEqual(x, y)
        # Handle None
        elif x is None:
            self.testcase.assertIsNone(y)
        # Default case
        else:
            # raise NotImplementedError(f'PythonCorrector cannot compare {type(x)=}')
            self.testcase.assertEqual(x, y)

    @gracefully_check
    def check_equals_on_grid(self, f, args, values):
        for x, y in zip(args, values, strict=False):
            self.testcase.assertEqual(f(x), y)

    @gracefully_check
    def check_module_is_imported(self, module_name, teacher):
        # TODO mad: handle from_imports
        self.testcase.assertIn(module_name, teacher.story['user_input']['imports']['imports'])

    def check_stdout_equals(self, x):
        # NOTE mad: no need for the `gracefully_check` decorator as the `check_equal` method is used
        stdout = self.stdout.getvalue()
        stdout = stdout.strip().replace('\n', '')
        x = x.strip().replace('\n', '')
        self.check_equal(stdout, x)

    def check_nb_lines(self): ...

    # TODO: mad
    # check if nb lines consistent with sections
    # check if nb lines ==

    def correct(self) -> bool:
        ...
        # TODO: ideally the checks would be performed there
        # answer_nature = answer['nature'].split('_')
        # return self.correction

_pc = PythonCorrector('')

# _________________public_checks_________________


# _________________private_checks_________________

Legacy

Codex_ original work

# _________________foreground_script_________________


# _________________corrector_________________
from io import StringIO
from pprint import pprint
import sys
import typing
import unittest
class PythonCorrector:
    def __init__(self, background_script):
        self.background_script = background_script
        self.last_error: str | None = None
        self.testcase = unittest.TestCase()
        self.check_counter = 0
        self.correction = {}
        # TODO: move this in the teacher class
        # Create a StringIO object to capture output
        self.original_stdout = sys.stdout
        # self.stdout = StringIO()
        # sys.stdout = self.stdout
        self.stdout = sys.stdout

        # if pyodide
        # todo : write surcouche pour data get

        print(sys.platform)

        #     print(codex_id)
        #     # load the data-background-script with js
        #     self.background_script = ''
        #     # js.globals.data_background_script

        #     self.background_script = document.activeElement.id

    def __del__(self):
        sys.stdout = sys.__stdout__

    @staticmethod
    def gracefully_check(check_method):
        def wrapper(self, *args, **kwargs):
            try:
                check_method(self, *args, **kwargs)
                success, msg = 1, ''
            except AssertionError as e:
                success, msg = 0, str(e)

            self.check_counter += 1
            self.correction[self.check_counter] = (success, msg)
            return success, msg

        return wrapper

    def check_coincide(self, x: typing.Any) -> None:
        namespace = {}
        exec(self.background_script, namespace)
        # TODO : SAME SETUP FOR ALL BECAUSE OF AUTOFORMAT
        x_name = f'{x=}'.split('=')[0]
        self.check_equal(x, namespace[x_name])

    @gracefully_check
    def check_equal(self, x: typing.Any, y: typing.Any) -> None:
        # First check if types match
        self.testcase.assertIsInstance(x, type(y))
        # Booleans
        if isinstance(x, bool):
            self.testcase.assertEqual(x, y)
        # Strings
        if isinstance(x, str):
            self.testcase.assertEqual(x, y)
        # Numbers
        elif isinstance(x, int):
            self.testcase.assertEqual(x, y)
        elif isinstance(x, float):
            self.testcase.assertAlmostEqual(x, y)
        # Sequences
        elif isinstance(x, list):
            self.testcase.assertListEqual(x, y)
        elif isinstance(x, tuple):
            self.testcase.assertTupleEqual(x, y)
        elif isinstance(x, set):
            self.testcase.assertSetEqual(x, y)
        # Dictionaries
        elif isinstance(x, dict):
            self.testcase.assertDictEqual(x, y)
        # Handle None
        elif x is None:
            self.testcase.assertIsNone(y)
        # Default case
        else:
            # raise NotImplementedError(f'PythonCorrector cannot compare {type(x)=}')
            self.testcase.assertEqual(x, y)

    @gracefully_check
    def check_equals_on_grid(self, f, args, values):
        for x, y in zip(args, values, strict=False):
            self.testcase.assertEqual(f(x), y)

    @gracefully_check
    def check_module_is_imported(self, module_name, teacher):
        # TODO mad: handle from_imports
        self.testcase.assertIn(module_name, teacher.story['user_input']['imports']['imports'])

    def check_stdout_equals(self, x):
        # NOTE mad: no need for the `gracefully_check` decorator as the `check_equal` method is used
        stdout = self.stdout.getvalue()
        stdout = stdout.strip().replace('\n', '')
        x = x.strip().replace('\n', '')
        self.check_equal(stdout, x)

    def check_nb_lines(self): ...

    # TODO: mad
    # check if nb lines consistent with sections
    # check if nb lines ==

    def correct(self) -> bool:
        ...
        # TODO: ideally the checks would be performed there
        # answer_nature = answer['nature'].split('_')
        # return self.correction

_pc = PythonCorrector('')

# _________________public_checks_________________


# _________________private_checks_________________

Sandbox - Interactive Python Execution

Python built-in libraries

Basic example

🔴 Inputs Not working

matplotlib (+ pandas) examples

Simple matplotlib example

pandas-only example

pandas + matplotlib time series example

bokeh interactive plots

🟢 Working example (for pm-codex)

🔴 Working example (for jupyter)

🟡 Deprecation warning (for pm-codex)

pm-codex integration tests

With height_in_px

With script_path