amerand / candid Goto Github PK

Companion Analysis and Non-Detection in Interferometric Data

Python 100.00%

candid's Introduction

I am an 🔭astronomer🔭 keen on sharing my experience and expertise via my codes! I thouroughly check my implementations for accuracy, but these are still research codes, and they may contain bugs: If you use my codes, please give me feedback (report issues, request features, etc). The codes below have reference publications, and have been used in more than 80 refereed publications as of mid-2023.

SPIPS models spectroscopic, photometric and interferometric time series observations of Cepheids pulsating stars. It is used to derive fundamental parameters, and interpolated light curves. It is described in Mérand et al. A&A 581-80 (2015). ✨Highlight✨ It has been used to create a PCA templates of light curve to re-analise independetly some of the SH0ES data in Javanmardi, Mérand et al. ApJ 911-12 (2021) to measure the Hubble constant
CANDID aims at detecting faint companions around stars in interferometric data, co-developped with Alexandre Gallenne (who hosts a version with a GUI). It gives you an upper limt for non-detection. The algorithms are described in Gallenne, Mérand et al. A&A 579-68 (2015). ✨Highlight✨ CANDID has been used in 60 refereed publications in less than 8 years!
PMOIRED is an advanced tool to model interferometric data and help astrophyscal interpretation. The algorithms are described in Mérand, SPIE Volume 12183, id. 121831N (2022) ✨Highlight✨ It was used to disprove the presence of a black hole in the HR6819 system, as published in Frost et al. A&A 659-L3 (2022)

candid's People

Contributors

Stargazers

Watchers

Forkers

mikeireland agallenne tomasstolker

candid's Issues

Print statement breaks with single wavelength value

Hello,
First, thank you for this package -- we used it during the commissioning of JWST AMI mode and it has been very useful.
With the latest update (version 1.0.7, #1a3151c0fdc1484f11fe06c6608d9e94df13c3cc) I find that candid.py lines 1295-1296 break the fitting routine when the EFF_WAVE is a single float, specifically:

1295     print(' | EFF_BAND/gradient(EFF_WAVE)', hdu.header['INSNAME'], '~',
-> 1296         '%.3f'%np.mean(hdu.data['EFF_BAND']/np.gradient(hdu.data['EFF_WAVE'])))

ValueError: Shape of array too small to calculate a numerical gradient, at least (edge_order + 1) elements are required.

The issue here is taking e.g. np.gradient([3.8e-6]) since EFF_WAVE can be a single value for one channel/filter. Could this print statement be altered to accommodate that?
Thank you.

Gradient calculation error for SPHERE-SAM data

Dear CANDID builders,

I downloaded the latest version of CANDID a few days ago in order to process some SPHERE-SAM data (after extracting the observables using AMICAL). Note that I cross-checked the consistency and validity of the oifits files created with other dedicated tools from JMMC.

However, when I open my oifits file with CANDID, I have the following error:

cal1cal2 = candid.Open(file) 
   ...:                                                                                                                                                   
 | loading file /xxx/xxx/
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-9-c04a41cc609e> in <module>
      1 file = '/xxx/xxx/SPHERE_K1_CAL1CAL2.oifits'
----> 2 cal1cal2 = candid.Open(file)

~/Coding/CANDID-master/candid/__init__.py in __init__(self, filename, rmin, rmax, reducePoly, wlOffset, alpha, v2bias, instruments, largeCP)
   1137             self.titleFilename = os.path.basename(filename)
   1138             self._initOiData()
-> 1139             self._loadOifitsData(filename, reducePoly=reducePoly, largeCP=largeCP)
   1140 
   1141         print(' | compute aux data for companion injection')

~/Coding/CANDID-master/candid/__init__.py in _loadOifitsData(self, filename, reducePoly, largeCP)
   1297                 if type(hdu.data['EFF_WAVE'])==np.ndarray:
   1298                     print(' | EFF_BAND/gradient(EFF_WAVE)', hdu.header['INSNAME'], '~',
-> 1299                         '%.3f'%np.mean(hdu.data['EFF_BAND']/np.gradient(hdu.data['EFF_WAVE'])))
   1300                 else:
   1301                     # -- nothing to do

<__array_function__ internals> in gradient(*args, **kwargs)

~/opt/anaconda3/envs/candid_env/lib/python3.6/site-packages/numpy/lib/function_base.py in gradient(f, *varargs, **kwargs)
   1046         if f.shape[axis] < edge_order + 1:
   1047             raise ValueError(
-> 1048                 "Shape of array too small to calculate a numerical gradient, "
   1049                 "at least (edge_order + 1) elements are required.")
   1050         # result allocation

ValueError: Shape of array too small to calculate a numerical gradient, at least (edge_order + 1) elements are required.

It seems to be due to the EFF_BAND keyword as SPHERE data contains only one wavelength. The gradient computation on the wavelength table therefore crashes and prevents from using CANDID on SPHERE-SAM data...

Thanks a lot for your help,
All the best,
Faustine

amerand / candid Goto Github PK

candid's Introduction

candid's People

Contributors

Stargazers

Watchers

Forkers

candid's Issues

Print statement breaks with single wavelength value

Gradient calculation error for SPHERE-SAM data

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent