Comments (4)
I opened the derived pickle file. The values 'SOI_5_T_tra_tot' and 'SOI_5_T_tra_full' contain a first NaN value at the index = 0. Therefore, the piece of code from deriver.derive
doing this:
ind_good = []
for i,name in enumerate(names):
if (name in derived_samples) and isinstance(derived_samples[name], np.ndarray) and not any(np.isnan(derived_samples[name])) and not all(np.array(derived_samples[name])==0):
ind_good.append(i)
seems to be filtering the durations for this companion. I guess this is intended, but I'm wondering what is the cause of having NaN values and why we would ignore this derived parameter when only having one NaN value out of more than 30,000 as my current scenario.
Kind regards,
Martín.
from allesfitter.
Hi again @martindevora,
Ha, thanks for pushing allesfitter to its limits and providing all those detailed insights!
Hm... the question is, why is the first entry a NaN? There shouldn't be any NaN in that array if things go normal. It particularly puzzles me that it is the first one, too. This smells like something with the sampler went wrong. Have you had any new insights into this system since then? I'd be especially curious about re-running it from a different random state and checking the outcome.
Generally speaking, our philosophy in allesfitter is to err on the site of caution, given so many possible use cases and things-that-may-go-wrong. Since this is such an edge case, I would put this issue on ice for now.
(Again, thanks for your patience while I was moving from MIT to ESA this summer.)
from allesfitter.
Hi again here in some additional thread : )
I don't remember which one, but this happened again two weeks ago with some other system I was exploring too. It seemed that the last candidate (as happened in the case I uploaded to this issue) was missing the duration due to the same problem (first value from the samples is NaN). So I believe this is not an isolated problem. I must recognize that I'm trying to fit very shallow transits, but still, I'm providing the initial priors for the duration.
PS: Good luck with your new position in the ESA!
from allesfitter.
Hi @martindevora,
Ok, twice is too often :) In v1.2.7 (pip-installable by this evening) I have replaced the not any()
with an not all()
, and am using np.nanpercentile functions to allow computing the desired medians and lower/upper limits. I tested it on your example above and it ran well.
(I am still not happy that there is a NaN in the first position of the samples, but it is hard to diagnose why that is.)
Please feel free to reopen this if you find a continuing/related issue in v1.2.7 and beyond.
from allesfitter.
Related Issues (20)
- Quick setup
- [Bug?] allesfitter.allesclass(object_dir) fails HOT 2
- Customising plots - model kwargs not working
- Plotting bug for RM-Effect
- Implement enhanced CHEOPS detrending
- Implement alternative light curve models HOT 1
- ValueError while running "allesfitter.show_initial_guess('allesfit')" HOT 4
- Installation error in 'ellc' package HOT 5
- Unable to apply logarithmic Limb darkening law HOT 3
- Error when running the estimate_noise_out_of_transit function HOT 2
- New release HOT 1
- Question: SNR value of final fit HOT 2
- Phase Curves tutorial HOT 1
- allesclass bug
- how to fit RV stellar variation for different RV datasets HOT 4
- Star spots tutorial not available HOT 1
- How to make custom plot of phased RV or lightcurves
- np.float deprecated HOT 2
- Results folder just partly populated
- Error on startup: np.int deprecated
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from allesfitter.