Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    glitch
    @glitch
    Hi @LouisJenkinsCS can you give us some details about what was going on when you ran into this error?
    • What version of Arkouda are you running on the server
    • What version of HDF5 is installed
    • What is the internal data structure of the files you were trying to read? (i.e. where they generated by Arkouda or generated somewhere else?)
      Thanks!
    5 replies
    Michael Merrill
    @mhmerrill
    We will be discussing multi-dim array views today on the Arkouda weekly call.
    Hope @bmcdonald3 and @glitch can make the call today along with anyone else who would like to be in on the conversation!
    Louis Jenkins
    @LouisJenkinsCS
    @glitch I responded in thread form btw
    glitch
    @glitch
    @LouisJenkinsCS gotcha, thanks!
    Michael Merrill
    @mhmerrill
    ATTENTION: We are no longer generating Arkouda Documentation on the Read-The-Docs site, we are only generating documentation onto GitHub Pages
    glitch
    @glitch
    Here's a direct link to the docs: https://bears-r-us.github.io/arkouda/
    You can also find the link in the README.md :thumbsup:
    Michael Merrill
    @mhmerrill
    We are NOT having an Arkouda Weekly Call today
    Happy Holidays!
    Michael Merrill
    @mhmerrill
    sorry I missed having the weekly call today
    Michael Merrill
    @mhmerrill
    Today on the Arkouda Weekly Call we will discuss the code that handles binary operators and its structure. We will also take a minute to discuss the next PRs we are going to merge.
    Michael Merrill
    @mhmerrill
    I am cancelling todays Arkouda Weekly Call
    glitch
    @glitch
    got it
    Zhihui Du
    @zhihuidu
    @glitch The BFS pull request code has been updated based on your suggestions. If you have time, please have a look and let us know your comments. If you have time, we may have a zoom discussion about it. Thanks!
    Michael Merrill
    @mhmerrill
    Today's Arkouda weekly call @bmcdonald3 will talk about his PR #1034 to refactor the binary operators
    Michael Merrill
    @mhmerrill
    I am cancelling todays Arkouda Weekly Call
    Michael Merrill
    @mhmerrill
    I am cancelling today's Arkouda Weekly Call ... again... unless someone wants a meeting?
    Zhihui Du
    @zhihuidu

    For the latest Arkouda version. I have the following error

    /home/z/zd4/Mike/arkouda//dep/checkRE2.chpl:3: In function 'main':
    /home/z/zd4/Mike/arkouda//dep/checkRE2.chpl:9: error: Cannot find module or enum 'Regex'
    make: * [check-re2] Error 1

    I have installed regex using conda.
    How to fix this checking problem?

    pierce314159
    @pierce314159
    you need to set the CHPL_RE2 env variable and rebuild chapel
    export CHPL_RE2=bundled
    Zhihui Du
    @zhihuidu
    Thanks!
    login-1-84 chapel >: export CHPL_RE2=bundled
    login-1-85 chapel >: make
    ....
    make[3]: Nothing to be done for `re2'.
    .....
    It seems I missed something else?
    I need to
    make clean
    first?
    Must I build it from clean? If I rebuild the chapel, it will need a long time.
    pierce314159
    @pierce314159
    if arkouda is still failing to make, then try make clean and then make
    Brad Chamberlain
    @bradcray
    I would expect you not to have to re-build if re2 is already built.
    What version of Chapel are you using? (chpl --version)
    Zhihui Du
    @zhihuidu
    chpl version 1.24.0 pre-release (8ca0d00)
    I have make clean and make
    Brad Chamberlain
    @bradcray
    I think the problem is actually in the checkRE2.chpl file itself. There’s a Makefile variable you can set to skip this check, and I would recommend doing that and seeing if Arkouda proper builds for you.
    make “ARKOUDA_SKIP_CHECK_DEPS=true from the arkouda directory.
    Zhihui Du
    @zhihuidu
    Thanks and I will try it.
    Brad Chamberlain
    @bradcray
    @pierce314159 : I see you worked on checkRE2.chpl last. The issue that I think I’m seeing is that use statements are resolved early in the compiler, such that both Regex and Regexp will both attempt to be parsed by the compiler regardless of version number.
    pierce314159
    @pierce314159
    ohhhhh
    Brad Chamberlain
    @bradcray
    Arkouda proper deals with this by bringing in ArkoudaRegexCompat.chpl, which publicly uses one module or the other depending on version number. (another approach would’ve been to add a module named Regex to the lt-125 directory that publicly used Regexp).
    We’ve recently been talking about what could be done to make uses get processed more lazily (or other things to be processed more aggressively) so that the code would’ve worked as you’d hoped. But that’s a ways off yet, I’m afraid.
    pierce314159
    @pierce314159
    okay yeah that makes sense, thanks Brad!
    Zhihui Du
    @zhihuidu
    @bradcray Yes, make ARKOUDA_SKIP_CHECK_DEPS=true can pass the checking and now everything seems good.
    Brad Chamberlain
    @bradcray
    Great!
    @pierce314159 : I added a comment to #1032 here: https://github.com/Bears-R-Us/arkouda/pull/1032/files#r802103781 but probably it deserves an issue to make sure it doesn’t get lost. Can you take care of that?
    pierce314159
    @pierce314159
    yeah I just created: Bears-R-Us/arkouda#1079
    Brad Chamberlain
    @bradcray
    :+1:
    pierce314159
    @pierce314159
    @zhihuidu, the issue you were having should be resolved by Bears-R-Us/arkouda#1084. You should be able to make without skipping checks on the latest version of arkouda master
    Zhihui Du
    @zhihuidu
    @pierce314159 yes, I have tested it and the new version works! Thanks!
    Zhihui Du
    @zhihuidu
    @bradcray Hi, Brad, our tests show that on a shared memory multicore computer, it will take a much longer time if we run a program with two locales compared with one locale. I wonder what chaple will do when it cannot have multiple physical resources to execute a program?
    Brad Chamberlain
    @bradcray
    Hi @zhihuidu — Sorry not to have noticed this earlier. I believe that you're correct that, today by default, Chapel processes assume they own the full compute node, and so will (for example) create a thread per core and pin those threads to the cores. When both processes do this, it can hurt performance of course.
    There are some advanced features that can be used to try and carve up a compute node (like a shared memory system like this) between multiple Chapel processes, and we've had some success with them, but it's still not optimal.