Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Feb 20 18:31
    peterbell10 commented #43
  • Feb 20 15:58
    joaosferreira edited #43
  • Feb 20 15:58
    joaosferreira opened #43
  • Feb 20 15:34
    joaosferreira opened #42
  • Feb 20 07:51
    prasunanand commented #41
  • Feb 20 07:50
    prasunanand assigned #41
  • Feb 20 07:50
    prasunanand opened #41
  • Feb 17 10:59
    hameerabbasi commented #227
  • Feb 17 10:58

    hameerabbasi on master

    fix some problems in notebooks … (compare)

  • Feb 17 10:58
    hameerabbasi closed #227
  • Feb 13 14:01
    sangyx commented #40
  • Feb 13 13:50
    hameerabbasi commented #40
  • Feb 13 13:48
    sangyx commented #40
  • Feb 13 13:45
    sangyx opened #227
  • Feb 13 13:20
    hameerabbasi commented #40
  • Feb 13 13:19

    hameerabbasi on master

    Fix NumPy backend issue, always… Merge branch 'master' of github… (compare)

  • Feb 13 12:36
    hameerabbasi commented #40
  • Feb 13 12:22
    sangyx commented #40
  • Feb 13 10:19
    hameerabbasi labeled #40
  • Feb 13 10:18
    hameerabbasi commented #40
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] > So the x.diffs[x] is the first order derivative of x.array?
More accurately, it’s the first order derivative of x wrt itself.
joaosferreira
@joaosferreira
@quansightai_twitter Which different variable names would you use ( e.g., x1 and x2)?
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @joaosferreira Those names work.
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] I also saw your question on GitHub... this is only inside the registered diff functions. Elsewhere, standard NumPy notation applies.
Shiv Kumar
@championballer
@quansightai_twitter Were the minutes of the meeting on unumpy documented? Could really help me get started. :)
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @championballer We don’t keep them unfortunately. But there is documentation, and you can ask questions anytime. :)
Sayandip Halder
@sayandip18
register_diff(np.cbrt, lambda x: x[1] / (3 (x[0] * (2 / 3))))
The cbrt and sqrt functions are different. If we need to differentiate ( say x^5) then how is it supposed to be computed?
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @sayandip18 It goes to np.pow, which computes the general f(x) ** g(x), which you can Google the derivative of.
Sayandip Halder
@sayandip18
Why are sqrt and cbrt different? Isn't np.pow supposed to take care of them too?
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @sayandip18 Because NumPy has them, we register the derivative for them. :man-shrugging::skin-tone-3:
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] And also because they’re different functions, essentially.
[Hameer Abbasi, Quansight] Since we’re doing a dictionary lookup, we need to hit all three cases, as sqrt and cbrt go to C, not through pow.
Ishdutt Trivedi
@ishdutt
Hello Everyone! I was searching for the projects and found Uarray (Expanding it's overall coverage )interesting..So form where can I start?
Quansight
@quansightai_twitter
[Unknown, Quansight]
Issue opened by sk-ip
Quansight-Labs/udiff#1
The project directory structure does not look good, compared to the standard library structure.
for making packages for PyPI the directory structure should be something like this
project-root
├──src/
   └──package_name/
      └──__init__.py
├──tests/
   └──package_name/
├──.coveragerc
├──.gitignore
├──LICENSE
├──README.md
├──setup.cfg
└──setup.py
Sudhanshu kumar
@sk-ip
Hello I am Sudhanshu, an undergraduate student.
I made an issue, above on GitHub. I would like to work on this if the above mentioned directory structure is fine.
Shubham Mishra
@grapheo12
Hello. I am Shubham Mishra. I am interested in this project. Are any slots still available for GSoC?
Quansight
@quansightai_twitter
[Unknown, Quansight]
Pull request opened by sk-ip
Quansight-Labs/udiff#2
references Quansight-Labs/udiff#1 closes Quansight-Labs/udiff#1
made the directory structure according to standard python library structure.
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] Hi @grapheo12. We already have a lot of students for udiff, so your best bet is uarray or unumpy.
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @ishdutt Hi! Adding a single multimethod in a PR would be a good way to start,
Quansight
@quansightai_twitter
[Unknown, Quansight]
Pull request opened by prasunanand
Quansight-Labs/udiff#3
Reviewers: @hameerabbasi
Quansight
@quansightai_twitter
[Unknown, Quansight]
*1 new commit pushed to master*
Quansight-Labs/uarray@173d9b8 - fix some problems in notebooks (#227)
[Unknown, Quansight]
Pull request merged by hameerabbasi
Quansight-Labs/uarray#227
[Unknown, Quansight]
Issue closed by hameerabbasi
Quansight-Labs/udiff#1
[Unknown, Quansight]
*1 new commit pushed to master*
Quansight-Labs/udiff@119c969 - Improve the directory structure. (#2)
[Unknown, Quansight]
Pull request merged by hameerabbasi
Quansight-Labs/udiff#2
Quansight
@quansightai_twitter
[Unknown, Quansight]
*1 new commit pushed to master*
Quansight-Labs/udiff@25db161 - Add README, Contributing Guidelines and conda environment (#3)
[Unknown, Quansight]
Pull request merged by hameerabbasi
Quansight-Labs/udiff#3
Mukul Mehta
@mukul-mehta
Hello Everybody! I'm Mukul Mehta and I'm interested to work on this project. My first choice was udiff but since there's already lot of people, I'd like to contribute to uarray. Any issues I could start with?
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] Hi, @mukul-mehta! Do you have your own idea you'd like to bring along or the one on the ideas page?
[Hameer Abbasi, Quansight] You can also try unumpy which is kind of dull but straightforward once you get the hang of it.
Hameer Abbasi
@hameerabbasi
Just so that we're aware here, @sangyx and @sk-ip are the only ones who've made PRs, and are the only ones who qualify for a slot thus far. It'd be good to get a PR in if anyone wants to qualify.
Quansight
@quansightai_twitter
[Unknown, Quansight]
Issue opened by joaosferreira
Quansight-Labs/udiff#4
As stated in the GSoC project ideas, add some test cases for the functions computing derivatives in _builtin_diffs.py.
Quansight
@quansightai_twitter
[Unknown, Quansight]
Pull request opened by joaosferreira
Quansight-Labs/udiff#5
This is intended to resolve issue Quansight-Labs/udiff#4 with the same name.
sakshamb2113
@sakshamb2113
Hi I am Saksham . I am new to open source and want to contribute to uarray. Can you help me get started?
Vyom16
@Vyom16
Hello Community !! I am Vyom Goel from a student from Computer Science Department,IIT Mandi . I found this organisation quite intriguing . I am very new to Open Source .So, can anyone please guide me how to contribute here.
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] Hi @sakshamb2113 and @Vyom16. The GSoC ideas page has a guide on how to contribute to open source, it is pretty general and also applies to this project. You should read that.
Quansight
@quansightai_twitter

[Unknown, Quansight]
Pull request opened by sangyx
Quansight-Labs/udiff#6
Hi, I found a bug in np.positive and np.negative. The diff calculate in the source is incorrect:

register_diff(np.positive, lambda x: +x[1])
register_diff(np.negative, lambda x: -x[1])

If we suppose that x = [1, -1], then the y = np.positive(x)=[1, 1]. We can see that y = x if x > 0 and y=-x if x < 0. So the diff of y should be y=sign(x)=[1, -1]

joaosferreira
@joaosferreira
Hi, is there an implementation of np.array_equal in unumpy?
Quansight
@quansightai_twitter
[Prasun Anand, Quansight] Don't think so.
joaosferreira
@joaosferreira
Should an issue be open for this? I also notice that values like np.e and np.NaN are not defined in unumpy.
Quansight
@quansightai_twitter
[Unknown, Quansight]
Issue opened by prasunanand
Quansight-Labs/unumpy#41
Quansight
@quansightai_twitter
[Unknown, Quansight]
*1 new commit pushed to master*
Quansight-Labs/udiff@e23163c - add np.sign & np.abs (#6)
[Unknown, Quansight]
Pull request merged by hameerabbasi
Quansight-Labs/udiff#6
Quansight
@quansightai_twitter
[Hameer Abbasi, Quansight] @joaosferreira Those seem to be good issues and simple PRs. One would just create a Python float with the value of pi and e.
joaosferreira
@joaosferreira
Thanks, I will try to do these today.
Quansight
@quansightai_twitter
[Unknown, Quansight]
Issue opened by joaosferreira
Quansight-Labs/unumpy#42
These constants are missing in the project.
Quansight
@quansightai_twitter

[Unknown, Quansight]
Pull request opened by joaosferreira
Quansight-Labs/unumpy#43
This resolves Quansight-Labs/unumpy#42.

It seems a very simple PR. If there is more appropriate way of adding constants please let me know.

chetan9191
@chetan9191
I want to take part in this project .