<div dir="ltr"><div class="gmail_default" style="font-family:arial,helvetica,sans-serif">Hi Stefan, Edmund,</div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><br></div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif">Thanks - and sorry that the update didn't work.  I think I understand why that is, as I took out some "requirements" for the conda package.  I will update that package to put the pyepics requirement back in, especially as it really needs to be updated since Python 3.7.6 broke pyepics (no, really... long story, and we have a work-around).  I do think that doing a "conda update --all" might have worked, but re-installing is probably the safest bet.</div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><br></div><div class="gmail_default" style="font-family:arial,helvetica,sans-serif"> </div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Mar 2, 2020 at 3:54 AM Edmund Welter <<a href="mailto:edmund.welter@desy.de">edmund.welter@desy.de</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
  
    
  
  <div bgcolor="#FFFFFF">
    <p>Dear Matt, Stefan,</p>
    <p>I fear it is the same on my Ubuntu 16.04 desktop. After updating
      I get the error messages shown in the attached text file when i
      try to start larch in the terminal.</p>
    <p>Cheers,</p>
    <p>Edmund</p>
    <p><br>
    </p>
    <div>On 02.03.20 10:36, Mangold, Stefan
      (IPS) wrote:<br>
    </div>
    <blockquote type="cite">
      
      <div>Dear all,</div>
      <div><br>
      </div>
      <div>tried the upgrade under anaconda3, on MacOs10.12.6.
        after the upgrade nothing worked any more </div>
      <div><br>
      </div>
      <div>did the following steps:</div>
      <div> </div>
      <div>conda update --all</div>
      <div>conda install -yc GSECARS xraylarch</div>
      <div>larch -m</div>
      <div><br>
      </div>
      <div>and</div>
      <div>pip install pyshortcuts==1.4</div>
      <div>larch -m</div>
      <div><br>
      </div>
      <div>nothing worked anymore. It worked again after a
        complete delete and re-install of the actual version of
        anaconda3</div>
      <div><br>
      </div>
      <div>Best regards</div>
      <div><br>
      </div>
      <div>Stefan Mangold</div>
      <br>
      <div>
        <blockquote type="cite">
          <div>Am 28.02.2020 um 21:53 schrieb Matt Newville
            <<a href="mailto:newville@cars.uchicago.edu" target="_blank">newville@cars.uchicago.edu</a>>:</div>
          <br>
          <div>
            <div dir="ltr">
              <div class="gmail_default" style="font-family:arial,helvetica,sans-serif">Hi
                Everyone,<br>
              </div>
              <div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><br>
                </div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><span style="font-family:Arial,Helvetica,sans-serif">Larch 0.9.47 is now available, with
                    installers and source code at  </span><a href="https://xraypy.github.io/xraylarch/installation.html" style="font-family:Arial,Helvetica,sans-serif" target="_blank">https://xraypy.github.io/xraylarch/installation.html</a>. 
                   <span style="font-family:Arial,Helvetica,sans-serif">For python users, there is a</span><font face="arial, helvetica, sans-serif"> plain
                    python package available on PyPI and conda packages
                    for A</font><span style="font-family:Arial,Helvetica,sans-serif">naconda Python.</span>  See the
                  installation docs for more details. </div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><span style="font-family:Arial,Helvetica,sans-serif"><br>
                  </span></div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif"><span style="font-family:Arial,Helvetica,sans-serif">There have been several improvements and
                    bug fixes, especially for the XAS Viewer application
                    and for XRF modeling in the nearly six months since
                    the last release.  In particular, there have been
                    two improvements to basic XAFS and XANES data
                    processing, both based on user reports and
                    comparisons to older versions of Ifeffit/Athena and
                    give a noticeable change in XAFS and XANES
                    processing.</span><br>
                </div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif">
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">First, the ranges used in by the pre_edge()
                    function for finding the edge step for normalization
                    are now better determined from the actual data range
                    rather than simply being hard-wired numbers.  These
                    improvements were long over-due and give noticeably
                    better default results for XANES data, especially
                    for relatively low-energy edges such as S and Cl K
                    edges.  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">When reading Athena Project files (say, to
                    import into XAS Viewer), the pre-edge and
                    normalization ranges from the Athena Project file
                    will be preserved.  When reading in new raw data, or
                    if you select the "Use Default Setting" button on
                    the Normalization Panel for any group in XAS Viewer,
                    the newer defaults will be used.   You can always
                    alter these values, but in playing around with this
                    with a range of datasets, the new defaults seem to
                    give a noticeable improvement in almost all cases
                    and rarely bad.<br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">Second, as a few users have pointed out or
                    gently hinted at over many months, there were
                    sometimes significant differences in the background
                    removals between classic Autobk/Ifeffit/Athena and
                    Larch, with Larch sometimes being noticeably and
                    inexplicably worse. I believe this involved two
                    different problems.  One was introduced a while back
                    when implementing an estimate of delta_chi - the
                    variance in chi due to the background subtraction.
                    This estimate is important, but I botched some of
                    the configurations of the number of knots, fit
                    range, and Rbkg. The other problem was that "spline
                    clamps" were just done too differently in Larch and
                    Ifeffit/Athena.</div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">I believe this is now working much better:
                    the background results are much more consistent, and
                    do not occasionally get "very bad".  They also
                    happen to be generally closer to
                    Autobk/Ifeffit/Athena, and perhaps slightly better
                    because the fit range in R-space is now more
                    consistently determined (instead of wandering +/- a
                    few R data points around Rbkg where the misfit will
                    often be the largest). In addition, `delta_chi`
                    (never calculated in Ifeffit/Athena) is now also
                    more consistent.  One consequence of this change is
                    that a very small change in Rbkg (of say 0.01 to
                    0.05 Ang) may actually give no difference at all in
                    mu0(E) or in chi(k).  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">I bring these changes up because I think
                    they will be noticeable.  I think they are both
                    improvements, but let me know if you find cases for
                    which you think are now made worse.   Possibly
                    related: one thing that I definitely noticed in
                    going through several example data sets was that I
                    tended to favor a k-weight of 2 instead of 1 for
                    background subtraction -- so much so that it seemed
                    like this might be a better default.  I did not
                    change this default yet, but if you have a strong
                    opinion on this, that might be a good topic for
                    discussion here.</div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                </div>
                <div class="gmail_default" style="font-family:arial,helvetica,sans-serif">
                  <div style="font-family:Arial,Helvetica,sans-serif">
                    <div>There are some documentation
                      improvements, but this is an ongoing process and
                      never complete.  It is also one area where help
                      and feedback would greatly be appreciated.  If you
                      or your students have time to work through the
                      larch examples and/or documentation and make
                      improvements or even suggestions for improvements
                      in readability or completeness, it would be
                      greatly appreciated.</div>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif">Thanks,</div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><br>
                  </div>
                  <div style="font-family:Arial,Helvetica,sans-serif"><span class="gmail_default" style="font-family:arial,helvetica,sans-serif"></span></div>
                </div>
                <div dir="ltr" style="font-family:arial,helvetica,sans-serif">--Matt Newville </div>
              </div>
            </div>
            _______________________________________________<br>
            Ifeffit mailing list<br>
            <a href="mailto:Ifeffit@millenia.cars.aps.anl.gov" target="_blank">Ifeffit@millenia.cars.aps.anl.gov</a><br>
            <a href="http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit" target="_blank">http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit</a><br>
            Unsubscribe:
            <a href="http://millenia.cars.aps.anl.gov/mailman/options/ifeffit" target="_blank">http://millenia.cars.aps.anl.gov/mailman/options/ifeffit</a><br>
          </div>
        </blockquote>
      </div>
      <br>
      <br>
      <fieldset></fieldset>
      <pre>_______________________________________________
Ifeffit mailing list
<a href="mailto:Ifeffit@millenia.cars.aps.anl.gov" target="_blank">Ifeffit@millenia.cars.aps.anl.gov</a>
<a href="http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit" target="_blank">http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit</a>
Unsubscribe: <a href="http://millenia.cars.aps.anl.gov/mailman/options/ifeffit" target="_blank">http://millenia.cars.aps.anl.gov/mailman/options/ifeffit</a>
</pre>
    </blockquote>
  </div>

_______________________________________________<br>
Ifeffit mailing list<br>
<a href="mailto:Ifeffit@millenia.cars.aps.anl.gov" target="_blank">Ifeffit@millenia.cars.aps.anl.gov</a><br>
<a href="http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit" rel="noreferrer" target="_blank">http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit</a><br>
Unsubscribe: <a href="http://millenia.cars.aps.anl.gov/mailman/options/ifeffit" rel="noreferrer" target="_blank">http://millenia.cars.aps.anl.gov/mailman/options/ifeffit</a><br>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature">--Matt Newville <newville at <a href="http://cars.uchicago.edu" target="_blank">cars.uchicago.edu</a>> 630-252-0431<br></div>