Scripting Tutorial

CyxWiz provides two ways to run Python code:

• Console - Interactive REPL for quick commands
• Script Editor - Write and save .cyx or .py files

Opening the Console

1. Go to View > Console (or press Ctrl+`)
2. The Console panel appears at the bottom of the screen

Running Commands

>>> 2 + 2
4

>>> import pycyxwiz as cx
>>> cx.get_version()
'0.1.0'

>>> print("Hello, CyxWiz!")
Hello, CyxWiz!

>>> import pycyxwiz as cx
>>> print("CUDA:", cx.cuda_available())
CUDA: True
>>> print("OpenCL:", cx.opencl_available())
OpenCL: False

>>> devices = cx.get_available_devices()
>>> for d in devices:
...     print(f"{d.name}: {d.memory_total // (1024**3)} GB")
NVIDIA RTX 3080: 10 GB

1. Go to File > New Script (or press Ctrl+N)
2. Choose script type:
   • Python (.py) - Standard Python file
   • CyxWiz Script (.cyx) - Notebook-style cells

CyxWiz scripts (.cyx) use a cell-based format similar to Jupyter notebooks:

%%code
# This is a code cell
import pycyxwiz as cx
print("Hello from cell 1")

%%code
# This is another code cell
data = [1, 2, 3, 4, 5]
print(sum(data))

%%markdown
# Documentation Cell
This is markdown text that won't be executed.

Create a new script and write:

%%code
# Cell 1: Import and setup
import pycyxwiz as cx
import numpy as np

print("CyxWiz Version:", cx.get_version())
print("NumPy Version:", np.__version__)

%%code
# Cell 2: Create a tensor
data = np.random.randn(3, 4).astype(np.float32)
tensor = cx.Tensor.from_numpy(data)

print("Tensor shape:", tensor.shape())
print("Tensor dtype:", tensor.get_data_type())

%%code
# Cell 3: Basic operations
t1 = cx.Tensor.random([4, 4])
t2 = cx.Tensor.random([4, 4])

result = t1 + t2  # Element-wise addition
print("Sum result shape:", result.shape())

%%code
# Cell 4: Neural network layer
dense = cx.Dense(10, 5)  # 10 inputs, 5 outputs
input_tensor = cx.Tensor.random([2, 10])  # Batch of 2, 10 features

output = dense.forward(input_tensor)
print("Output shape:", output.shape())  # [2, 5]

%%code
import matplotlib.pyplot as plt
import numpy as np

# Generate data
x = np.linspace(0, 10, 100)
y = np.sin(x)

# Create plot
plt.figure(figsize=(8, 4))
plt.plot(x, y, 'b-', label='sin(x)')
plt.xlabel('x')
plt.ylabel('y')
plt.title('Sine Wave')
plt.legend()
plt.grid(True)
plt.show()  # Plot appears inline

The plot renders directly below the cell!

%%code
# Setup
import pycyxwiz as cx
import numpy as np

# Create synthetic data
np.random.seed(42)
X = np.random.randn(100, 4).astype(np.float32)
y = np.random.randint(0, 3, 100)  # 3 classes

# Convert to tensors
X_tensor = cx.Tensor.from_numpy(X)

%%code
# Build model
layer1 = cx.Dense(4, 16)
relu = cx.ReLU()
layer2 = cx.Dense(16, 3)
softmax = cx.Softmax()

# Forward pass
def forward(x):
    x = layer1.forward(x)
    x = relu.forward(x)
    x = layer2.forward(x)
    x = softmax.forward(x)
    return x

%%code
# Test forward pass
output = forward(X_tensor)
print("Output shape:", output.shape())  # [100, 3]
print("First prediction:", output.to_numpy()[0])

• Using the Console for interactive Python
• Creating .cyx notebook-style scripts
• Running cells individually or all at once
• Creating plots that render inline
• Basic model building with pycyxwiz