Scikit-RF

1.11.0 · active · verified Thu Apr 16

Scikit-RF is an object-oriented Python library for microwave engineering. It provides tools for network analysis, circuit simulation, calibration, and visualization of S-parameters and other RF components. The current version is 1.11.0, and it maintains an active, though not strictly time-bound, release cadence.

Common errors

```
AttributeError: module 'skrf' has no attribute 'Network'
```
cause Attempting to access `Network` directly as `skrf.Network` after using `from skrf import Network` or similar, or vice-versa, or due to a partial/incorrect import.

fix The canonical way to use `Network` is `import skrf as rf; ntwk = rf.Network(...)`. Ensure you are consistently using `rf.Network` after `import skrf as rf` or `from skrf import Network` if you prefer direct import.
```
ValueError: Frequency unit must be one of ['hz', 'khz', 'mhz', 'ghz', 'thz']
```
cause An invalid string was provided for the `unit` argument when creating a `Frequency` object.

fix Correct the `unit` string to one of the accepted values: 'hz', 'khz', 'mhz', 'ghz', or 'thz'. For example: `rf.Frequency(1, 10, 101, 'ghz')`.
```
FileNotFoundError: [Errno 2] No such file or directory: 'my_file.s2p'
```
cause The specified Touchstone file path is incorrect, or the file does not exist at the given location.

fix Double-check the file path. Ensure the file exists, the path is absolute or correct relative to your script's execution directory, and that you have read permissions. Use `os.path.exists()` for debugging.
```
IndexError: index 0 is out of bounds for axis 1 with size 0
```
cause This often occurs when trying to access elements of an S-parameter matrix (e.g., `ntwk.s[:,0,0]`) on a `Network` object that hasn't been properly initialized with S-parameter data or has zero ports/frequencies.

fix Ensure the `Network` object has valid S-parameter data and dimensions. If creating from scratch, provide `s` (S-parameters array) and `frequency` to the `Network` constructor. If loading, verify the source file's integrity.

Warnings

gotcha Scikit-RF handles frequency units internally, but users often confuse input units (Hz, kHz, MHz, GHz) leading to incorrect calculations or plots. Always be explicit.
Fix: When creating a `Frequency` object, ensure the `unit` argument matches your numerical input (e.g., `rf.Frequency(1, 10, 101, 'ghz')` for GHz input). For existing `Network` objects, access frequency in desired units via properties like `ntwk.f`, `ntwk.f_khz`, `ntwk.f_mhz`, `ntwk.f_ghz`.
breaking Prior to version 0.16, the `skrf.network.Network` class and other components had different constructors or attribute names. Code written for older versions might fail due to these changes.
Fix: Consult the scikit-rf changelog and upgrade guide if migrating from very old versions. Common breaking changes include `Network` constructor arguments or internal attribute restructuring. Ensure your code aligns with the `1.x` series API.
gotcha When loading Touchstone files, scikit-rf automatically determines the number of ports. If the file is malformed or unusual, this can lead to incorrect network dimensions or parsing errors.
Fix: Validate your Touchstone file format. If issues persist, try loading the file into a text editor to confirm its structure and headers. For programmatic inspection, `rf.Network(file='path/to/file.s2p').nports` can confirm the detected port count.
gotcha Default plotting styles might not always be ideal for publication or specific analysis. Matplotlib's default may hide details or look unpolished.
Fix: Utilize Matplotlib's extensive customization options. After calling `ntwk.plot_s_db(...)`, you can use `plt.xlabel()`, `plt.ylabel()`, `plt.title()`, `plt.grid(True)`, `plt.ylim()`, `plt.style.use('seaborn-v0_8-darkgrid')` (or other styles) to enhance plots. Scikit-rf plots return Matplotlib axes, allowing direct manipulation.

Install

pip install scikit-rf Install stable version

Imports

Network
wrong:
```
from scikit_rf import Network
```
correct:
```
import skrf as rf
ntwk = rf.Network(...)
```
The common convention is to import `skrf` as `rf` and access classes through it. `Network` is the core class representing a multi-port network.
Frequency
wrong:
```
from skrf.frequency import Frequency
```
correct:
```
import skrf as rf
freq = rf.Frequency(...)
```
While functional, direct import from submodules is less common than accessing via the `skrf` namespace.
Circuit
wrong:
```
from skrf.circuit import Circuit
```
correct:
```
import skrf as rf
cir = rf.Circuit(...)
```
Similar to `Frequency`, the preferred way is through the top-level `skrf` namespace.

Quickstart

This quickstart demonstrates how to create a simple 2-port `Network` object in scikit-rf, populate it with dummy S-parameter data, and then plot its S21 magnitude in dB. It highlights the use of `Frequency` and `Network` classes and the common `plot_s_db` method.

import skrf as rf
import matplotlib.pyplot as plt

# Create a dummy 2-port network for demonstration
f = rf.Frequency(1, 10, 101, 'ghz') # 1-10 GHz, 101 points
s = rf.Touchstone(f, name='my_network')

# Populate with some dummy S-parameters (e.g., ideal thru) for 2-port
s.s[:,0,0] = 0 # S11
s.s[:,1,1] = 0 # S22
s.s[:,0,1] = 1 # S12
s.s[:,1,0] = 1 # S21

# Create a Network object from the Touchstone data
ntwk = rf.Network(s=s.s, frequency=f, name='Dummy Thru')

# Plot S21 magnitude in dB
ntwk.plot_s_db(m=1, n=0)
plt.title(f'{ntwk.name} S21 Magnitude')
plt.xlabel('Frequency (GHz)')
plt.ylabel('S21 Magnitude (dB)')
# plt.show() # Uncomment to display plot

print(f"Network name: {ntwk.name}")
print(f"Network frequency range: {ntwk.frequency.f_ghz[0]:.2f} GHz to {ntwk.frequency.f_ghz[-1]:.2f} GHz")

view raw JSON →