From j.loveday at sussex.ac.uk Thu Nov 1 11:24:48 2018 From: j.loveday at sussex.ac.uk (Jonathan Loveday) Date: Thu, 1 Nov 2018 15:24:48 +0000 Subject: [AstroPy] astropy.modeling.fitting.LevMarLSQFitter Message-ID: <861A9058-ECD2-4844-8F7B-DE8CD7F32DE7@sussex.ac.uk> Hi astropy.modelling developers A couple of points regarding this class: 1. The documentation http://docs.astropy.org/en/stable/api/astropy.modeling.fitting.LevMarLSQFitter.html#astropy.modeling.fitting.LevMarLSQFitter states: weights : array (optional) Weights for fitting. For data with Gaussian uncertainties, the weights should be 1/sigma. This should of course read "the weights should be 1/sigma^2? (unless the code is squaring the weights, which would be very confusing). 2. Is there an easy away to obtain the fitted chi^2, other than calculating it oneself from the returned fitting function? Presumably it must be calculated internally. Thanks, Jon Jon Loveday Astronomy Centre, University of Sussex, Falmer, Brighton, BN1 9QH, UK J.Loveday at sussex.ac.uk +44 (0) 1273 877719 From ejensen1 at swarthmore.edu Thu Nov 1 11:55:04 2018 From: ejensen1 at swarthmore.edu (Eric L. N. Jensen) Date: Thu, 1 Nov 2018 11:55:04 -0400 Subject: [AstroPy] astropy.modeling.fitting.LevMarLSQFitter In-Reply-To: <861A9058-ECD2-4844-8F7B-DE8CD7F32DE7@sussex.ac.uk> References: <861A9058-ECD2-4844-8F7B-DE8CD7F32DE7@sussex.ac.uk> Message-ID: <052E1848-FA8A-457D-90CE-07EC70165769@swarthmore.edu> Hi Jon, > On Nov 1, 2018, at 11:24 AM, Jonathan Loveday wrote: > > > 2. Is there an easy away to obtain the fitted chi^2, other than calculating it oneself from the returned fitting function? Presumably it must be calculated internally. This requires one more step to compute the chi^2, but it gets you pretty close from what the returned info is providing: https://stackoverflow.com/questions/14854339/in-scipy-how-and-why-does-curve-fit-calculate-the-covariance-of-the-parameter-es/17508924#17508924 Eric -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2354 bytes Desc: not available URL: From indiajoe at gmail.com Thu Nov 1 12:48:42 2018 From: indiajoe at gmail.com (Joe Ninan) Date: Thu, 1 Nov 2018 12:48:42 -0400 Subject: [AstroPy] astropy.modeling.fitting.LevMarLSQFitter In-Reply-To: <861A9058-ECD2-4844-8F7B-DE8CD7F32DE7@sussex.ac.uk> References: <861A9058-ECD2-4844-8F7B-DE8CD7F32DE7@sussex.ac.uk> Message-ID: Hi Jon, On Thu, 1 Nov 2018 at 11:44, Jonathan Loveday wrote: > Hi astropy.modelling developers > > A couple of points regarding this class: > > 1. The documentation > http://docs.astropy.org/en/stable/api/astropy.modeling.fitting.LevMarLSQFitter.html#astropy.modeling.fitting.LevMarLSQFitter > states: > > weights : array (optional) > Weights for fitting. For data with Gaussian uncertainties, the weights > should be 1/sigma. > > This should of course read "the weights should be 1/sigma^2? (unless the > code is squaring the weights, which would be very confusing). > > In the code, the multiplication of the residuals by 1/sigma is done before the final squaring and summing of the residuals for minimisation. This is also consistent with how one would implement the weight inside the residue function if they use scipy.optimize.leastsq So atleast in the current implementation the user should provide 1/sigma as weights of heteroscedastic data. -cheers joe _______________________________________________ > AstroPy mailing list > AstroPy at python.org > https://mail.python.org/mailman/listinfo/astropy > -- /--------------------------------------------------------------- "GNU/Linux: because a PC is a terrible thing to waste" - GNU Generation ************************************************ Joe Philip Ninan Postdoctoral Researcher 525 Davey Lab, Dept. of Astronomy & Astrophysics The Pennsylvania State University University Park, PA-16802 ------------------------------------------------------------ Website: https://indiajoe.gitlab.io/ My GnuPG Public Key: https://indiajoe.gitlab.io/files/JPN_public.key -------------- next part -------------- An HTML attachment was scrubbed... URL: From adrianmpw at gmail.com Mon Nov 5 09:31:12 2018 From: adrianmpw at gmail.com (Adrian Price-Whelan) Date: Mon, 5 Nov 2018 09:31:12 -0500 Subject: [AstroPy] 2019 Gaia Sprint - Santa Barbara (USA) Message-ID: Dear all -- Applications are now open for the 2019 Gaia Sprint in Santa Barbara: http://gaia.lol/ The idea behind the Gaia Sprints is to bring together people who are interested in timely investigation and use of the Gaia data, with no focus on scientific topic or subfield. Examples of projects started at the last sprint can be found in the wrap-up slides here: https://speakerdeck.com/dwhgg/2018-nyc-gaia-sprint-wrap-up-slides The next sprint will take place 25?29 March 2019 at the KITP in Santa Barbara and will be limited to ~40 local participants, but we anticipate having many simultaneous satellite sprints around the world (details to follow). Applications will close 10 December, 2018. If you have any questions, feel free to reach out to me or any of the organizers. Also feel free to share this message with any people or mailing lists who may be interested. See you in Santa Barbara! -- Adrian M. Price-Whelan Lyman Spitzer, Jr. Postdoctoral Fellow Princeton University http://adrn.github.io From thomas.robitaille at gmail.com Mon Nov 26 11:26:33 2018 From: thomas.robitaille at gmail.com (Thomas Robitaille) Date: Mon, 26 Nov 2018 16:26:33 +0000 Subject: [AstroPy] glue v0.14 released Message-ID: Hi everyone, We are happy to announce the release of *glue v0.14*! For anyone not familiar with this package, *glue* is a Python library and application for multidimensional and linked data exploration, which you can find out about at http://glueviz.org *Changes in this release* An overview of the main changes in this release is available here: http://docs.glueviz.org/en/stable/whatsnew/whatsnew.html The most significant changes are as follows: - The current active subset is now clearly shown in the toolbar at the top of the application window - Session files are now saved with relative paths to data files by default - Performance has been significantly improved in a number of places, and in particular the profile and histogram viewers will now no longer hang the application if calculating profiles or histograms for large datasets (they will do this in the background) - We now provide documentation about available plugins to glue as well as information about developing your own plugin packages . - The data layer of glue has been re-thought to make it possible to create data objects that are based on remote or non-cartesian data You can read about these changes in more detail at the link above. In addition, this release includes many bug fixes and improvements in usability, and we have also done a lot of internal work to facilitate the development of interfaces not based on Qt. *Installing/updating glue* As usual, we recommend installing glue using the Anaconda Python Distribution. To get the latest version of glue (v0.14), you will need to make sure you use the glueviz conda channel. If you are using the conda command, this means that you need to do: conda install -c glueviz glueviz=0.14 If you want to use the Anaconda launcher or navigator to install glue without using the command-line, take a look at the instructions on our website to find out how to get the latest version. We also provide instructions for other installation methods, including pip. Please let us know if you run into any issues installing or using glue - you can either let us know by email (by replying to this email off-list), by opening a GitHub issue , or you can join the glue slack channel and chat with us there. Thanks to everyone who contributed to this release! - Thomas Robitaille, on behalf of the glue team -------------- next part -------------- An HTML attachment was scrubbed... URL: