Mailman 3 June 2016 - SciPy-Dev

Review pull request #6285: BUG: stats: Inconsistency in the multivariate_normal docstring #6263
by Bhavika Tekwani June 19, 2016

June 19, 2016

Hey everyone, I just submitted a pull request attempting to close issue #6263 regarding the pdf and logpdf methods in multivariate_normal. As I understood - the bug was simply an absence of default parameters in pdf and logpdf methods within class multivariate_normal_gen. I've submitted this change along with a test for default values. Please see the PR here - https://github.com/scipy/scipy/pull/6285 I hope this is correct and I welcome guidance if it is not. Thank you, Bhavika T.

2 1

optimize: what should happen if objective functions return non-finite numbers?
by Andrew Nelson June 16, 2016

June 16, 2016

Consider the following example which raises an AssertionError: import numpy as np from scipy.optimize import minimize def func1(x): return np.nan x0 = np.array([1.3, 0.7, 0.8, 1.9, 1.2]) res = minimize(func1, x0, method='l-bfgs-b') assert(res.success = False) minimize simply returns the starting values of: res.x == x0. The reason I came up with this example is that unsanitised datasets sometimes contain nan or inf. Thus, if func1 was calculating chi2 and you were using minimize then the entire fit would appear to succeed (res.success is True), but the output would be garbage. Ok, so res.message is CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL, but it's not an clear indicator that something went wrong. A second example is: import numpy as np from scipy.optimize import curve_fit def func2(x, a, b, c): return a * np.exp(-b * x) + c def func3(x, a, b, c): return np.nan xdata = np.linspace(0, 4, 50) y = func2(xdata, 2.5, 1.3, 0.5) ydata = y + 0.2 * np.random.normal(size=len(xdata)) popt, pcov = curve_fit(func3, xdata, ydata) print(popt) Whilst there is a warning (OptimizeWarning: Covariance of the parameters could not be estimated) it's not a clear indicator that something has gone wrong. The behaviour one might expect in both examples could be to see a ValueError raised if there arenp.nan values returned from the objective function. I'm not totally sure of what to do if +/- np.inf is returned (-inf would be a very good global minimum). -- _____________________________________ Dr. Andrew Nelson _____________________________________

4 7

Re: [SciPy-Dev] SciPy-Dev Digest, Vol 152, Issue 6
by Matt Newville June 15, 2016

June 15, 2016

On Wed, Jun 15, 2016 at 7:00 AM, <scipy-dev-request(a)scipy.org> wrote: > Send SciPy-Dev mailing list submissions to > scipy-dev(a)scipy.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://mail.scipy.org/mailman/listinfo/scipy-dev > or, via email, send a message with subject or body 'help' to > scipy-dev-request(a)scipy.org > > You can reach the person managing the list at > scipy-dev-owner(a)scipy.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of SciPy-Dev digest..." > > > Today's Topics: > > 1. optimize: what should happen if objective functions return > non-finite numbers? (Andrew Nelson) > 2. Re: optimize: what should happen if objective functions > return non-finite numbers? (josef.pktd(a)gmail.com) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 15 Jun 2016 09:19:45 +1000 > From: Andrew Nelson <andyfaff(a)gmail.com> > To: SciPy Developers List <scipy-dev(a)scipy.org> > Subject: [SciPy-Dev] optimize: what should happen if objective > functions return non-finite numbers? > Message-ID: > < > CAAbtOZdiCO-_bcqF0UPy-Bn6SmmqZfcOPZmdEBAiQchbCpf_zA(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Consider the following example which raises an AssertionError: > > import numpy as np > from scipy.optimize import minimize > def func1(x): > return np.nan > x0 = np.array([1.3, 0.7, 0.8, 1.9, 1.2]) > res = minimize(func1, x0, method='l-bfgs-b') > assert(res.success = False) > > minimize simply returns the starting values of: res.x == x0. The reason I > came up with this example is that unsanitised datasets sometimes contain > nan or inf. Thus, if func1 was calculating chi2 and you were using minimize > then the entire fit would appear to succeed (res.success is True), but the > output would be garbage. Ok, so res.message is CONVERGENCE: > NORM_OF_PROJECTED_GRADIENT_<=_PGTOL, but it's not an clear indicator that > something went wrong. > A second example is: > > import numpy as np > from scipy.optimize import curve_fit > def func2(x, a, b, c): > return a * np.exp(-b * x) + c > > def func3(x, a, b, c): > return np.nan > > xdata = np.linspace(0, 4, 50) > y = func2(xdata, 2.5, 1.3, 0.5) > ydata = y + 0.2 * np.random.normal(size=len(xdata)) > > popt, pcov = curve_fit(func3, xdata, ydata) > print(popt) > > Whilst there is a warning (OptimizeWarning: Covariance of the parameters > could not be estimated) it's not a clear indicator that something has gone > wrong. > The behaviour one might expect in both examples could be to see a > ValueError raised if there arenp.nan values returned from the objective > function. I'm not totally sure of what to do if +/- np.inf is returned > (-inf would be a very good global minimum). > > > -- > _____________________________________ > Dr. Andrew Nelson > > > _____________________________________ >

1 0

PR: Deprecate lti cross-class properties/setters
by Felix Berkenkamp June 13, 2016

June 13, 2016

Hi all, I may be a bit late, but I just opened a PR that I would preferably still like to see go into 0.18. It deprecates historic properties/setters of lti systems, such as the ability to set the numerator of a StateSpace system. Internally these kind of operations are very hacky and numerically inaccurate, so I think it is best to deprecate them. https://github.com/scipy/scipy/pull/6224 Best, Felix

2 1

PR adding a cython interface for special
by Joshua Wilson June 12, 2016

June 12, 2016

Hello all, This is a request for comments on https://github.com/scipy/scipy/pull/6195 which adds a cython interface for scalar, typed versions of most of the functions in scipy.special. The documentation can be found here: https://circle-artifacts.com/gh/scipy/scipy/52/artifacts/0/home/ubuntu/scip… - Josh

1 0

Re: [SciPy-Dev] raise on boolean sparse.__neg__
by G Young June 8, 2016

June 8, 2016

Hi, I have had a PR <https://github.com/scipy/scipy/pull/6166> open for quite some time now that raises when we attempt to take the negative of a boolean sparse matrix. While this has been given the +1 by @perimosocordiae <https://github.com/perimosocordiae>, as this is a backwards-incompatible change, he wanted to make sure others were okay with this change before merging. The rationale for raising on a boolean sparse matrix is nicely explained by here <https://github.com/scipy/scipy/pull/6166#issuecomment-219628455>, and I wanted to send this out to the mailing list to see if other maintainers and developers agree with this logic. Thanks!

1 0

scipy 0.18 release
by Evgeni Burovski June 4, 2016

June 4, 2016

Hi, It's seems to be about time to start planning the release of scipy 0.18. I volunteer to do the legwork for this release. The proposed schedule would be: 20 June: rc1 27 June: rc2 (if necessary) 11 July: final release Thoughts? Here's the current contents of the milestone: https://github.com/scipy/scipy/milestones/0.18.0 Any help with addressing open issues and reviewing/merging open PRs would be most appreciated. Cheers, Evgeni

2 1