<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    [*Snip*]<br>
    <br>
       
    <blockquote type="cite">Group-by and related functionality is top on
      my list of priorities for astropy.table (in fact I see it every
      day on my google keep app...).  Join and merging are in master
      now.  In my tests the astropy table join is within a factor of 2
      to 3 in speed relative to pandas, so in most use cases it should
      be good enough.</blockquote>
    Join and merge would follow the pandas behaviour? Because that is
    one of its major assets, I believe - its elegant handling of missing
    data and misaligned indices etc.<br>
    Speed is a minor issue in my world, we're not often working with the
    data set sizes the quants are.<br>
    <br>
    <blockquote
cite="mid:CAMtEP6ypJOoJoneaX3qeiqDXctVmcpDVmEKx3xD=-+Vm1Bqwgg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div class="gmail_extra">
          <div class="gmail_quote">
            <div>It's probably worth pointing out to the community that
              it was not a lightly-taken decision to reject pandas for
              use as the base data storage container.  For the case of
              tables there is one show-stopper which is that pandas
              DataFrame does not support arbitrary multi-dimensional
              columns, i.e. column where each element is itself an N-d
              array.  These occur enough in astronomy and are supported
              by FITS and VO standards, so the astropy Table must be
              able to represent that.  The lack of support for table and
              column metadata is a smaller but still important issue.</div>
            <div><br>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    That is very interesting to hear. I had, in fact, been wondering
    what was the rationale, but it certainly makes sense.<br>
    I was under the impression, though, that pandas pretty much
    supported arbitrary objects as entries in its data structures. But I
    was apparently mistaken?<br>
    <br>
    <blockquote
cite="mid:CAMtEP6ypJOoJoneaX3qeiqDXctVmcpDVmEKx3xD=-+Vm1Bqwgg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div class="gmail_extra">
          <div class="gmail_quote">
            <div>Having said that, there is no question pandas has a ton
              of highly-efficient and useful machinery and we are
              working on ways to improve inter-operability.  This
              includes being able convert between Table and DataFrame
              easily.  Suggestions and (especially) pull requests
              welcome.</div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    An easy and near-seamless conversion between a DataFrame and an
    astropy table (as long as data type support allows for it, of
    course) would definitely be a great thing.<br>
    <br>
    <blockquote
cite="mid:CAMtEP6ypJOoJoneaX3qeiqDXctVmcpDVmEKx3xD=-+Vm1Bqwgg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div class="gmail_extra">
          <div class="gmail_quote">
            <div> </div>
            <blockquote class="gmail_quote" style="margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">
              <div dir="ltr">
                <div><br>
                </div>
                <div>I won't speculate about whether that's enough an
                  asset to warrant a dependency in astropy. I do agree
                  that lots of other pandas features don't translate as
                  well into astronomy use.</div>
                <div><br>
                </div>
              </div>
              <div class="gmail_extra">
                <div>
                  <div class="h5"><br>
                    <br>
                    <div class="gmail_quote">On Thu, Jun 20, 2013 at
                      12:34 PM, Erik Tollerud <span dir="ltr"><<a
                          moz-do-not-send="true"
                          href="mailto:erik.tollerud@gmail.com"
                          target="_blank">erik.tollerud@gmail.com</a>></span>
                      wrote:<br>
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div dir="ltr">
                          <div>I'm of mixed minds about traits UI
                            because once you know it you can make great
                            GUIs with it, but I've spent a lot of time
                            troubleshooting people's python
                            installations to get traits to work.  That
                            is, in general it can be tricky to get
                            installed because of all the dependencies.
                             Maybe this has improved recently with
                            Enthought's Canopy (or other new python
                            distros), but that's been my past
                            experience.  </div>
                          <div><br>
                          </div>
                          <div>More generally, the view in the astropy
                            core package is that we don't want to put
                            GUIs in the core because GUIs always carry
                            lots of dependencies, which we don't want to
                            be forced to deal with.  But part of the
                            whole reason for affiliated packages was to
                            get around this, so we're happy to see
                            GUI-based affiliated packages.</div>
                          <div><br>
                          </div>
                          <div><br>
                          </div>
                          <div>As for Pandas, to be totally honest, I
                            don't see a huge amount to be gained from
                            adding a Pandas dependency Astropy.  It's
                            honestly not clear what it gives the
                            astronomy community that numpy does not
                            already have.  The following quote from the
                            Pandas web site has guided me to that
                            conclusion: "<em
style="line-height:22.390625px;text-align:justify;font-size:medium;font-family:Arial,sans-serif;margin:0px;padding:0px">pandas</em><span
style="line-height:22.390625px;text-align:justify;font-size:medium;font-family:Arial,sans-serif"> helps
                              fill this gap, enabling you to carry out
                              your entire data analysis workflow in
                              Python without having to switch to a more
                              domain specific language like R."</span></div>
                          <div><span
style="line-height:22.390625px;text-align:justify;font-size:medium;font-family:Arial,sans-serif"><br>
                            </span></div>
                          <div>I have been carrying out my entire data
                            analysis workflow for some time now in
                            python without using Pandas.  It looks to me
                            like Pandas is a tool that was written by
                            and for statisticians who use R.  While we
                            can take lessons from this, it's not clear
                            we get much out of it in an astronomy
                            context. For example, how would it make
                            astropy's NDData, Quantity, or Table better
                            to use a Pandas DataFrame vs. a numpy array?
                            Most of what we are doing is
                            building astronomy-convenient interfaces,
                            and I'm not sure what Pandas adds there, at
                            the cost of a pretty heavy-weight
                            dependency.</div>
                          <div><br>
                          </div>
                          <div>It could just be that I don't know enough
                            about Pandas, though.  So if someone who
                            knows Pandas better can speak to this, I'm
                            all ears.</div>
                          <div>  </div>
                          <div><br>
                          </div>
                        </div>
                        <div class="gmail_extra">
                          <div>
                            <div><br>
                              <br>
                              <div class="gmail_quote">On Tue, Jun 18,
                                2013 at 3:35 PM, Thøger Rivera-Thorsen <span
                                  dir="ltr"><<a
                                    moz-do-not-send="true"
                                    href="mailto:trive@astro.su.se"
                                    target="_blank">trive@astro.su.se</a>></span>
                                wrote:<br>
                                <blockquote class="gmail_quote"
                                  style="margin:0 0 0
                                  .8ex;border-left:1px #ccc
                                  solid;padding-left:1ex">
                                  <div text="#000000" bgcolor="#FFFFFF">
                                    <div>Pandas is a part of the
                                      newly-defined SciPy stack, after
                                      all, so that would be part of any
                                      science-oriented distribution
                                      worth its salt. In fact, I think
                                      it could be a good idea for
                                      astropy in general to use under
                                      the hood, but again, could clash
                                      with the philosophy of the project
                                      and possibly also
                                      maintainabillity.<br>
                                      <br>
                                      As for offering my code or just my
                                      experience, I'll have to square it
                                      with my supervisor first, and I
                                      also think it depends on what
                                      direction the project in question
                                      will take. I'm positive about the
                                      idea (which is why I wrote in the
                                      first place), but supervisor might
                                      think it is a better idea to
                                      actually get my paper in the
                                      project wrapped up before sending
                                      the code out there. Will get back
                                      about that one!<br>
                                      <br>
                                      /Emil
                                      <div>
                                        <div><br>
                                          <br>
                                          <br>
                                          <br>
                                          <br>
                                          On 2013-06-18 20:53, Slavin,
                                          Jonathan wrote:<br>
                                        </div>
                                      </div>
                                    </div>
                                    <blockquote type="cite">
                                      <div>
                                        <div><font
                                            face="verdana,sans-serif">Hi
                                            Emil,</font>
                                          <div><font
                                              face="verdana,sans-serif"><br>
                                            </font></div>
                                          <div><font
                                              face="verdana,sans-serif">That
                                              looks very nice!  I don't
                                              see Pandas as a big issue
                                              in terms of dependencies.
                                               I don't know that much
                                              about traits, etc.  My
                                              thought about the gui was
                                              just based on my
                                              experience with
                                              matplotlib, and the fact
                                              that it is widely used --
                                              though I would agree that
                                              too many dependencies can
                                              be a deterrent to people
                                              using something.  Are you
                                              offering your code as a
                                              starting point for the
                                              project?  It strikes me
                                              that many have gotten some
                                              sort of fitting package to
                                              a point of personal
                                              usability but no one has
                                              the
                                              time/interest/motivation
                                              to make a more generally
                                              usable package.</font></div>
                                          <div><font
                                              face="verdana,sans-serif"><br>
                                            </font></div>
                                          <div><font
                                              face="verdana,sans-serif">Jon</font></div>
                                          <div><font
                                              face="verdana,sans-serif"><br>
                                            </font>
                                            <div class="gmail_quote">On
                                              Tue, Jun 18, 2013 at 2:34
                                              PM, <span dir="ltr"><<a
                                                  moz-do-not-send="true"
href="mailto:astropy-request@scipy.org" target="_blank">astropy-request@scipy.org</a>></span>
                                              wrote:<br>
                                              <blockquote
                                                class="gmail_quote"
                                                style="margin:0 0 0
                                                .8ex;border-left:1px
                                                #ccc
                                                solid;padding-left:1ex">
                                                <div
                                                  style="overflow:hidden">Date:
                                                  <span><span>Tue, 18
                                                      Jun 2013 20:39:55</span></span>
                                                  +0200<br>
                                                  From: Th?ger
                                                  Rivera-Thorsen <<a
moz-do-not-send="true" href="mailto:thoger.emil@gmail.com"
                                                    target="_blank">thoger.emil@gmail.com</a>><br>
                                                  Subject: Re: [AstroPy]
                                                  ESA Summer of Code in
                                                  Space 2013<br>
                                                  To: <a
                                                    moz-do-not-send="true"
href="mailto:astropy@scipy.org" target="_blank">astropy@scipy.org</a><br>
                                                  Message-ID: <<a
                                                    moz-do-not-send="true"
href="mailto:51C0A97B.8090703@gmail.com" target="_blank">51C0A97B.8090703@gmail.com</a>><br>
                                                  Content-Type:
                                                  text/plain;
                                                  charset="iso-8859-1"<br>
                                                  <br>
                                                  I have been working on
                                                  a fitting GUI for a
                                                  while, although it is
                                                  made<br>
                                                  with a specific task
                                                  in mind.<br>
                                                  However, it is not
                                                  based on Matplotlib
                                                  but on
                                                  Traits/Traitsui/Chaco
                                                  and<br>
                                                  Pandas. It is made for
                                                  a specific projhect
                                                  I'm working and as
                                                  such not<br>
                                                  yet usable for more
                                                  general cases, but it
                                                  could be a starting
                                                  point, if<br>
                                                  the dependencies don't
                                                  conflict with astropy
                                                  politics.<br>
                                                  <br>
                                                  Especially, I am happy
                                                  about the choice of
                                                  Pandas for managing a
                                                  quite<br>
                                                  complex data structure
                                                  (the fitted and/or
                                                  guessed values of an
                                                  arbitrary<br>
                                                  number of transitions
                                                  for an arbitrary
                                                  number of rows or
                                                  collapsed rows<br>
                                                  of a spatially
                                                  resolved spectrum) of
                                                  a), but also with the
                                                  Traits-based<br>
                                                  interactive interface
                                                  to build complex line
                                                  profiles from single<br>
                                                  gaussians, good for
                                                  fitting-by-eye and
                                                  giving good initial
                                                  guesses for<br>
                                                  fitting of complex
                                                  line profiles. It
                                                  hooks directly up to a
                                                  wrapper I've<br>
                                                  made for lmfit, but
                                                  given the modularity,
                                                  it should be
                                                  relatively easy<br>
                                                  to change to other
                                                  backends.<br>
                                                  <br>
                                                  It's still a
                                                  work-in-progress, but
                                                  there are some
                                                  screenshots here:<br>
                                                  <a
                                                    moz-do-not-send="true"
href="http://flic.kr/s/aHsjGaEMGg" target="_blank">http://flic.kr/s/aHsjGaEMGg</a>
                                                  .<br>
                                                  I know the choice and
                                                  number of dependencies
                                                  may be prohibitive but
                                                  it<br>
                                                  saved a lot of work on
                                                  the GUI, and Pandas
                                                  means the difference
                                                  between<br>
                                                  sanity and madness
                                                  when it comes to
                                                  keeping track of so
                                                  many parameters.<br>
                                                  <br>
                                                  Cheers,<br>
                                                  Emil<br>
                                                </div>
                                              </blockquote>
                                            </div>
                                            <br>
                                            <br>
                                            <br clear="all">
                                            <div>
                                              <div dir="ltr">________________________________________________________<br>
                                                Jonathan D.
                                                Slavin                
                                                Harvard-Smithsonian CfA<br>
                                                <a
                                                  moz-do-not-send="true"
href="mailto:jslavin@cfa.harvard.edu" target="_blank">jslavin@cfa.harvard.edu</a>   

                                                   60 Garden Street, MS
                                                83<br>
                                                phone: <a
                                                  moz-do-not-send="true"
href="tel:%28617%29%20496-7981" value="+16174967981" target="_blank">(617)
                                                  496-7981</a>      
                                                Cambridge, MA 02138-1516<br>
                                                fax: <a
                                                  moz-do-not-send="true"
href="tel:%28617%29%20496-7577" value="+16174967577" target="_blank">(617)
                                                  496-7577</a>   
                                                        USA<br>
________________________________________________________<br>
                                                <br>
                                              </div>
                                            </div>
                                          </div>
                                          <br>
                                          <fieldset></fieldset>
                                          <br>
                                        </div>
                                      </div>
                                      <pre>_______________________________________________
AstroPy mailing list
<a moz-do-not-send="true" href="mailto:AstroPy@scipy.org" target="_blank">AstroPy@scipy.org</a>
<a moz-do-not-send="true" href="http://mail.scipy.org/mailman/listinfo/astropy" target="_blank">http://mail.scipy.org/mailman/listinfo/astropy</a>
</pre>
                                    </blockquote>
                                    <br>
                                  </div>
                                  <br>
_______________________________________________<br>
                                  AstroPy mailing list<br>
                                  <a moz-do-not-send="true"
                                    href="mailto:AstroPy@scipy.org"
                                    target="_blank">AstroPy@scipy.org</a><br>
                                  <a moz-do-not-send="true"
                                    href="http://mail.scipy.org/mailman/listinfo/astropy"
                                    target="_blank">http://mail.scipy.org/mailman/listinfo/astropy</a><br>
                                  <br>
                                </blockquote>
                              </div>
                              <br>
                              <br clear="all">
                              <div><br>
                              </div>
                            </div>
                          </div>
                          <span><font color="#888888">-- <br>
                              Erik
                            </font></span></div>
                        <br>
                        _______________________________________________<br>
                        AstroPy mailing list<br>
                        <a moz-do-not-send="true"
                          href="mailto:AstroPy@scipy.org"
                          target="_blank">AstroPy@scipy.org</a><br>
                        <a moz-do-not-send="true"
                          href="http://mail.scipy.org/mailman/listinfo/astropy"
                          target="_blank">http://mail.scipy.org/mailman/listinfo/astropy</a><br>
                        <br>
                      </blockquote>
                    </div>
                    <br>
                    <br clear="all">
                    <div><br>
                    </div>
                    -- <br>
                  </div>
                </div>
                ************************************<span class="HOEnZb"><font
                    color="#888888"><br>
                    Chris Beaumont<br>
                    Graduate Student<br>
                    Institute for Astronomy<br>
                    University of Hawaii at Manoa<br>
                    2680 Woodlawn Drive<br>
                    Honolulu, HI 96822<br>
                    <a moz-do-not-send="true"
                      href="http://www.ifa.hawaii.edu/%7Ebeaumont"
                      target="_blank">www.ifa.hawaii.edu/~beaumont</a><br>
                    ************************************
                  </font></span></div>
              <br>
              _______________________________________________<br>
              AstroPy mailing list<br>
              <a moz-do-not-send="true" href="mailto:AstroPy@scipy.org">AstroPy@scipy.org</a><br>
              <a moz-do-not-send="true"
                href="http://mail.scipy.org/mailman/listinfo/astropy"
                target="_blank">http://mail.scipy.org/mailman/listinfo/astropy</a><br>
              <br>
            </blockquote>
          </div>
          <br>
        </div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
AstroPy mailing list
<a class="moz-txt-link-abbreviated" href="mailto:AstroPy@scipy.org">AstroPy@scipy.org</a>
<a class="moz-txt-link-freetext" href="http://mail.scipy.org/mailman/listinfo/astropy">http://mail.scipy.org/mailman/listinfo/astropy</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>