WICKED: Wiggle Corrector toolKit for NIRSpEc Data

Introduction

WiCKED (Wiggle Interference Corrector toolKit for NIRSpEc Data) is a python package designed to remove sinusoidal wiggles, also known as Moire patterns, that appear in NIRSpec IFS data. These patterns arise due to the undersampling of the Point Spread Function (PSF).

Overview

The Moire pattern can be modeled as a series of different sinusoidal waves. WiCKED uses a two different integrated spectrum templates a power-law and a second degree polynomial to model single-pixel spectra. The residual of the best-fit and the single-pixel spectrum is then fitted a series of sinusoidal waves plus a constant to effectively remove these wiggles.

To flag pixels affected by wiggles in the datacube, WiCKED calculates the Fourier Transfrom for the residual between the best-fit and the single-pixel spectrum and compares the mean amplitude at frequency were wiggles dominate (tipically f < 50 [1/micron]) and at longer wavelenghts. Pixels with a larger ratio between these two windows are flagged. The typical frequency of the wiggles is calculated based on the wiggles of the brightest pixel.

Usage Examples

A Detailed example on how to use WICKED is shown in the example notebook in the Example folder.

Installation

To install WICKED, you can pip install the repository on you local folder. There is no installation required for WiCKED.

pip install git+https://github.com/antoinedumontneira/WiCKED

Dependencies

WICKED requires the following Python libraries: - cap_mpfit - threadpoolctl - scipy - photutils - astropy

WICKED uses the Michelle Cappellari version of the mpfit library included in the mgefit version 5.0.15 available for installation HERE!. WICKED will automatically install mgefit 5.0.15 in the enviroment

Credits

I got the inspiration to use a polynomial fit to characterize the frequency trend of Wiggles from the code by M. Perna. You can find their original repository here. Their code fits a polynomial to the frequency of wiggles at different wavelengths for the central and brightest pixel in the datacube, and then uses this as a prior for the rest of the pixels. I use a similar approach, but the polynomial is also tuned based on a chi-square fit for each pixel, since wiggles in different pixels sometimes have different frequencies.

Citation

If you use WICKED, please cite our paper:

[WIggle Corrector Kit for NIRSpEc Data: WICKED] – [Dumont] et al. (2025)
https://ui.adsabs.harvard.edu/abs/2025arXiv250309697D/abstract

Contact

This python package is still under development. For any questions or feedback, please contact me via dumont (at) mpia.de