Mailman 3 July 2017 - NumPy-Discussion

mixed mode arithmetic
by Neal Becker July 11, 2023

July 11, 2023

I've been browsing the numpy source. I'm wondering about mixed-mode arithmetic on arrays. I believe the way numpy handles this is that it never does mixed arithmetic, but instead converts arrays to a common type. Arguably, that might be efficient for a mix of say, double and float. Maybe not. But for a mix of complex and a scalar type (say, CDouble * Double), it's clearly suboptimal in efficiency. So, do I understand this correctly? If so, is that something we should improve?

4 6

Invalid value encoutered : how to prevent numpy.where to do this?
by Eric Emsellem Feb. 18, 2023

Feb. 18, 2023

Dear all, I have a code using lots of "numpy.where" to make some constrained calculations as in: data = arange(10) result = np.where(data == 0, 0., 1./data) # or data1 = arange(10) data2 = arange(10)+1.0 result = np.where(data1 > data2, np.sqrt(data1-data2), np.sqrt(data2-data2)) which then produces warnings like: /usr/bin/ipython:1: RuntimeWarning: invalid value encountered in sqrt or for the first example: /usr/bin/ipython:1: RuntimeWarning: divide by zero encountered in divide How … [View More]

5 4

Re: [Numpy-discussion] example reading binary Fortran file
by Neil Martinsen-Burrell July 22, 2021

July 22, 2021

David Froger <david.froger.info <at> gmail.com> writes: > Hy,My question is about reading Fortran binary file (oh no this question > again...) I've posted this before, but I finally got it cleaned up for the Cookbook. For this purpose I use a subclass of file that has methods for reading unformatted Fortran data. See http://www.scipy.org/Cookbook/FortranIO/FortranFile. I'd gladly see this in numpy or scipy somewhere, but I'm not sure where it belongs. > program … [View More]makeArray > implicit none > integer,parameter:: nx=10,ny=20 > real(4),dimension(nx,ny):: ux,uy,p > integer :: i,j > open(11,file='uxuyp.bin',form='unformatted') > do i = 1,nx > do j = 1,ny > ux(i,j) = real(i*j) > uy(i,j) = real(i)/real(j) > p (i,j) = real(i) + real(j) > enddo > enddo > write(11) ux,uy > write(11) p > close(11) > end program makeArray When I run the above program compiled with gfortran on my Intel Mac, I can read it back with:: >>> import numpy as np >>> from fortranfile import FortranFile >>> f=FortranFile('uxuyp.bin', endian='<') >>> uxuy = f.readReals(prec='f') # 'f' for default reals >>> len(uxuy) 400 >>> ux = np.array(uxuy[:200]).reshape((20,10)).T >>> uy = np.array(uxuy[200:]).reshape((20,10)).T >>> p = f.readReals('f').reshape((20,10)).T >>> ux array([[ 1., 2., 3., 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20.], [ 2., 4., 6., 8., 10., 12., 14., 16., 18., 20., 22., 24., 26., 28., 30., 32., 34., 36., 38., 40.], [ 3., 6., 9., 12., 15., 18., 21., 24., 27., 30., 33., 36., 39., 42., 45., 48., 51., 54., 57., 60.], [ 4., 8., 12., 16., 20., 24., 28., 32., 36., 40., 44., 48., 52., 56., 60., 64., 68., 72., 76., 80.], [ 5., 10., 15., 20., 25., 30., 35., 40., 45., 50., 55., 60., 65., 70., 75., 80., 85., 90., 95., 100.], [ 6., 12., 18., 24., 30., 36., 42., 48., 54., 60., 66., 72., 78., 84., 90., 96., 102., 108., 114., 120.], [ 7.,Proxy-Connection: keep-alive Cache-Control: max-age=0 14., 21., 28., 35., 42., 49., 56., 63., 70., 77., 84., 91., 98., 105., 112., 119., 126., 133., 140.], [ 8., 16., 24., 32., 40., 48., 56., 64., 72., 80., 88., 96., 104., 112., 120., 128., 136., 144., 152., 160.], [ 9., 18., 27., 36., 45., 54., 63., 72., 81., 90., 99., 108., 117., 126., 135., 144., 153., 162., 171., 180.], [ 10., 20., 30., 40., 50., 60., 70., 80., 90., 100., 110., 120., 130., 140., 150., 160., 170., 180., 190., 200.]]) >>> uy array([[ 1. , 0.5 , 0.33333334, 0.25 , 0.2 , 0.16666667, 0.14285715, 0.125 , 0.11111111, 0.1 , 0.09090909, 0.08333334, 0.07692308, 0.07142857, 0.06666667, 0.0625 , 0.05882353, 0.05555556, 0.05263158, 0.05 ], [ 2. , 1. , 0.66666669, 0.5 , 0.40000001, 0.33333334, 0.2857143 , 0.25 , 0.22222222, 0.2 , 0.18181819, 0.16666667, 0.15384616, 0.14285715, 0.13333334, 0.125 , 0.11764706, 0.11111111, 0.10526316, 0.1 ], [ 3. , 1.5 , 1. , 0.75 , 0.60000002, 0.5 , 0.42857143, 0.375 , 0.33333334, 0.30000001, 0.27272728, 0.25 , 0.23076923, 0.21428572, 0.2 , 0.1875 , 0.17647059, 0.16666667, 0.15789473, 0.15000001], [ 4. , 2. , 1.33333337, 1. , 0.80000001, 0.66666669, 0.5714286 , 0.5 , 0.44444445, 0.40000001, 0.36363637, 0.33333334, 0.30769232, 0.2857143 , 0.26666668, 0.25 , 0.23529412, 0.22222222, 0.21052632, 0.2 ], [ 5. , 2.5 , 1.66666663, 1.25 , 1. , 0.83333331, 0.71428573, 0.625 , 0.55555558, 0.5 , 0.45454547, 0.41666666, 0.38461539, 0.35714287, 0.33333334, 0.3125 , 0.29411766, 0.27777779, 0.2631579 , 0.25 ], [ 6. , 3. , 2. , 1.5 , 1.20000005, 1. , 0.85714287, 0.75 , 0.66666669, 0.60000002, 0.54545456, 0.5 , 0.46153846, 0.42857143, 0.40000001, 0.375 , 0.35294119, 0.33333334, 0.31578946, 0.30000001], [ 7. , 3.5 , 2.33333325, 1.75 , 1.39999998, 1.16666663, 1. , 0.875 , 0.77777779, 0.69999999, 0.63636363, 0.58333331, 0.53846157, 0.5 , 0.46666667, 0.4375 , 0.41176471, 0.3888889 , 0.36842105, 0.34999999], [ 8. , 4. , 2.66666675, 2. , 1.60000002, 1.33333337, 1.14285719, 1. , 0.8888889 , 0.80000001, 0.72727275, 0.66666669, 0.61538464, 0.5714286 , 0.53333336, 0.5 , 0.47058824, 0.44444445, 0.42105263, 0.40000001], [ 9. , 4.5 , 3. , 2.25 , 1.79999995, 1.5 , 1.28571427, 1.125 , 1. , 0.89999998, 0.81818181, 0.75 , 0.69230771, 0.64285713, 0.60000002, 0.5625 , 0.52941179, 0.5 , 0.47368422, 0.44999999], [ 10. , 5. , 3.33333325, 2.5 , 2. , 1.66666663, 1.42857146, 1.25 , 1.11111116, 1. , 0.90909094, 0.83333331, 0.76923078, 0.71428573, 0.66666669, 0.625 , 0.58823532, 0.55555558, 0.52631581, 0.5 ]]) >>> p array([[ 2., 3., 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21.], [ 3., 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22.], [ 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23.], [ 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24.], [ 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25.], [ 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26.], [ 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26., 27.], [ 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28.], [ 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28., 29.], [ 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28., 29., 30.]]) Note that you have to provide the shape information for ux and uy because fortran writes them together as a stream of 400 numbers. -Neil [View Less]

10 10

Dropping support for Accelerate
by Ilhan Polat Sept. 15, 2017

Sept. 15, 2017

A few months ago, I had the innocent intention to wrap LDLt decomposition routines of LAPACK into SciPy but then I am made aware that the minimum required version of LAPACK/BLAS was due to Accelerate framework. Since then I've been following the core SciPy team and others' discussion on this issue. We have been exchanging opinions for quite a while now within various SciPy issues and PRs about the ever-increasing Accelerate-related issues and I've compiled a brief summary about the ongoing … [View More]

5 16

quantile() or percentile()
by Chun-Wei Yuan Aug. 14, 2017

Aug. 14, 2017

There's an ongoing effort to introduce quantile() into numpy. You'd use it just like percentile(), but would input your q value in probability space (0.5 for 50%): https://github.com/numpy/numpy/pull/9213 Since there's a great deal of overlap between these two functions, we'd like to solicit opinions on how to move forward on this. The current thinking is to tolerate the redundancy and keep both, using one as the engine for the other. I'm partial to having quantile because 1.) I prefer … [View More]

7 13

Changing MaskedArray.squeeze() to never return masked
by Eric Wieser Aug. 10, 2017

Aug. 10, 2017

When using ndarray.squeeze, a view is returned, which means you can do the follow (somewhat-contrived) operation: >>> def fill_contrived(a): a.squeeze()[...] = 2 return a >>> fill_contrived(np.array([1])) array(2) However, when tried with a masked array, this can fail, breaking liskov subsitution: >>> fill_contrived(np.ma.array([1], mask=[True])) MaskError: Cannot alter the masked element. This fails because squeeze breaks the contract of returning … [View More]

3 3

ENH: ratio function to mimic diff
by Joseph Fox-Rabinovitz July 29, 2017

July 29, 2017

I have created PR#9481 to introduce a `ratio` function that behaves very similarly to `diff`, except that it divides successive elements instead of subtracting them. It has some handling built in for zero division, as well as the ability to select between `/` and `//` operators. There is currently no masked version. Perhaps someone could suggest a simple mechanism for hooking np.ma.true_divide and np.ma.floor_divide in as the operators instead of the regular np.* versions. Please let me know … [View More]

3 4

NumPy steering councils members
by Ralf Gommers July 25, 2017

July 25, 2017

Hi all, It has been well over a year since we put together the governance structure and steering council ( https://docs.scipy.org/doc/numpy-dev/dev/governance/people.html#governance-…). We haven't reviewed the people on the steering council in that time. Based on the criteria for membership I would like to make the following suggestion (note, not discussed with everyone in private beforehand): Adding the following people to the steering council: - Eric Wieser - Marten van Kerkwijk - Stephan … [View More]

5 6

Slice nested arrays, How to
by Bob July 24, 2017

July 24, 2017

Hello, I created the following array by converting it from a nested list: a = np.array([np.array([ 17.56578416, 16.82712825, 16.57992292, 15.83534836]), np.array([ 17.9002445 , 17.35024876, 16.69733472, 15.78809856]), np.array([ 17.90086839, 17.64315136, 17.40653009, 17.26346787, 16.99901931, 16.87787178, 16.68278558, 16.56006419, 16.43672445]), np.array([ 17.91147242, 17.2770623 , 17.0320501 , 16.73729491, 16.4910479 ])], dtype=object) I … [View More]

3 2

Fwd: [NumFOCUS Projects] Important: 2017 NumFOCUS Summit & Sustainability Workshop
by Ralf Gommers July 21, 2017

July 21, 2017

Hi all, This is a public service announcement: for the first time NumFOCUS is organising a sustainability workshop, details below. NumPy is able to send two representatives; the NumPy steering committee chose Chuck and Nathaniel. I'll be present as well as NumFOCUS board member. Last year there was a NumFOCUS summit as well, but no sustainability workshop. It looks like going forward there will be one a year of these. If there's more interest than tickets next year we'll look at rotating the … [View More]NumPy representatives. Cheers, Ralf ---------- Forwarded message ---------- From: Christie Koehler <christie(a)numfocus.org> Date: Thu, Jun 29, 2017 at 10:52 AM Subject: [NumFOCUS Projects] Important: 2017 NumFOCUS Summit & Sustainability Workshop To: "projects(a)numfocus.org" <projects(a)numfocus.org> Dear Project Leads, Following up from our “save the date” note a few weeks ago, here is our formal invitation to the 2017 NumFOCUS Summit & Sustainability Workshop taking place in *Austin, Texas on 10-11 October, 2017*. All participants planning to attend should register <https://ti.to/numfocus-sustainability/2017-summit-workshop> no later than July 19th, 2017. New at this year’s Summit is our first ever Sustainability Workshop. The goal of the Workshop is to guide project leads and core contributors in identifying an appropriate sustainability plan for their project as well as the initial steps required to start implementing that plan. We would like to have at least 1-2 representatives from each NumFOCUS sponsored project attend the workshop. Members of the NumFOCUS Board of Directors, Advisory Council, and Sustainability Advisory Board who would like to attend are welcome to do so. Project Leads are tasked with selecting which project representative(s) should attend the Summit/Workshop, according to their governance practices. Once you’ve selected your reps, you may simply forward this email to them and ask them to register. You do not need to otherwise inform NumFOCUS of your choice prior. ------------------------------ N.B. — FYI, NumFOCUS is hoping to arrange an event for the Austin data science community to benefit from the presence of so many core maintainers of the scientific computing stack. This would take place immediately prior to the NumFOCUS Summit. No details are available yet — if we succeed in securing a venue, NumFOCUS staff will be in touch to invite your participation. We just wanted to give you a friendly heads-up as you consider your travel plans. Direct questions about this event to info(a)numfocus.org. ------------------------------ FAQ (Please review before registering): What am I committing to if I register for the Summit/Workshop? Those attending the Summit/Workshop should plan to attend all day on the 10th and 11th. You can view the tentative schedule for the Summit/Workshop here <https://numfocus.github.io/projects-director/project-briefs/sustainability-…>. Additionally, each project sending a representative to the Summit should commit to preparing and delivering a 5-minute lightning talk about the state of their project. NumFOCUS staff will be available to help with these presentations. Who do you recommend we send to represent our project? While we recognize that people have multiple roles, we recommend sending 1 person who can represent the technical aspects of your project (such as a person who develops or maintains code) and 1 person who can represent the community/business aspects (such as a person who works with the user/developer community). Representatives can be paid or unpaid contributors, but should be sufficiently invested and involved in your project such that they will both want to and are able to continue the sustainability work started at the Workshop. Can we send more than 2 reps from our project? Maybe. Our travel budget is based on two people from each project. That number is also what we are basing our venue, catering, and other logistical arrangements upon. Once we start collecting registrations, however, we may find that not all projects can send two representatives and/or that some projects can cover the cost of travel for one or more of their reps. So, a max of two registrations for each project will be available and thereafter anyone from your project wanting to attend will be given the option of being added to the waitlist. Because registrations are “first-come, first-served” you should make sure your first two choices of representative register before subsequent ones. Are you covering travel expenses for project reps? For those traveling from outside the United States, too? Yes. We have a budget for reimbursing participants for travel-related costs, even those needing to travel from outside the United States. To ensure the best use of these funds, we’ll ask you when you register if you need NumFOCUS to reimburse you for travel. For all participants, we will handle hotel reservations and payments. We will book hotel accommodations for all participants for three nights: 9-11 October. Please email summit-travel(a)numfocus.org if you need different arrangements. Those participants needing to be reimbursed for other travel-related expenses should generally expect to cover those expenses at the time they are incurred and submit them to NumFOCUS for prompt reimbursement after the Summit/Workshop. But if this will present a hardship for you, let us know and we’ll make other arrangements. For further details about what we’ll reimburse you for, read our travel reimbursement policy (coming soon). If you are traveling from outside of the United States and might need assistance with visa or other immigration-related issues, please let us know as soon as possible. What if I have other questions not answered here? Ask via this list, reply directly to Christie, or email summit-info(a)numfocus.org. Cheers, Christie -- Christie Koehler Projects Director, NumFOCUS christie(a)numfocus.org +1 415-317-7603 mobile/signal Need to meet with me? https://calendly.com/numfocus-christie -- You received this message because you are subscribed to the Google Groups "Fiscally Sponsored Project Representatives" group. To unsubscribe from this group and stop receiving emails from it, send an email to projects+unsubscribe(a)numfocus.org. To post to this group, send email to projects(a)numfocus.org. Visit this group at https://groups.google.com/a/numfocus.org/group/projects/ . To view this discussion on the web visit https://groups.google.com/a/ numfocus.org/d/msgid/projects/8EA01960-BE4A-4B5E-B1BD- 6A417243BC1B%40numfocus.org <https://groups.google.com/a/numfocus.org/d/msgid/projects/8EA01960-BE4A-4B5…> . [View Less]

1 0