Re: [Ifeffit] Larch-0.9.24 with Python2

Hi Matthias, Yes, that looks correct. I also fixed the code so that it can work with scipy 0.11 and later. I will work on changing the installation to check for minimal version numbers. On Wed, Nov 5, 2014 at 10:03 AM, Matthias Filez wrote:

Dear Matt,

We have upgraded to scipy 0.14.0 and this is what I get now when I run the script:

Read ASCII File: ['array_labels', 'attrs', 'column_labels', 'data', 'energy', 'filename', 'header', 'i0', 'mu'] After pre-edge: ['array_labels', 'attrs', 'column_labels', 'data', 'dmude', 'e0', 'edge_step', 'energy', 'filename', 'flat', 'header', 'i0', 'mu', 'nnorm', 'norm', 'norm1', 'norm2', 'norm_c0', 'norm_c1', 'norm_c2', 'norm_c3', 'nvict', 'post_edge', 'pre1', 'pre2', 'pre_edge', 'pre_offset', 'pre_slope'] After autobk: ['array_labels', 'attrs', 'autobk_details', 'bkg', 'chi', 'chie', 'column_labels', 'data', 'dmude', 'e0', 'edge_step', 'energy', 'filename', 'flat', 'header', 'i0', 'k', 'mu', 'nnorm', 'norm', 'norm1', 'norm2', 'norm_c0', 'norm_c1', 'norm_c2', 'norm_c3', 'nvict', 'post_edge', 'pre1', 'pre2', 'pre_edge', 'pre_offset', 'pre_slope'] [ 5.22014125e-01 4.68863165e-01 4.23387732e-01 3.84601353e-01 3.51517555e-01 3.23149862e-01 2.98511802e-01 2.76616899e-01 2.56478680e-01 2.37110672e-01 2.17526399e-01 1.96774248e-01 1.74202338e-01 1.49312809e-01 1.21645288e-01 9.07843838e-02 5.63993015e-02 1.85571728e-02 -2.03243717e-02 -5.75444687e-02 -9.14850299e-02 -1.18154030e-01 -1.36108084e-01 -1.45468821e-01 -1.47073697e-01 -1.50934997e-01 -1.54219769e-01 -1.56582731e-01 -1.55007074e-01 -1.50082523e-01 -1.35737843e-01 -1.13771275e-01 -8.25087559e-02 -4.62194044e-02 9.36626084e-04 4.42260206e-02 8.89877045e-02 1.27496716e-01 1.50717679e-01 1.52846156e-01 1.34118361e-01 1.02181582e-01 6.49568133e-02 3.22399885e-02 6.46113335e-03 -6.98123272e-03 -4.09343210e-03 1.83744265e-02 5.17624756e-02 8.65303295e-02 9.62396887e-02 7.69122752e-02 4.26727610e-02 7.49257701e-03 -1.81363351e-02 -3.96258643e-02 -6.15635768e-02 -7.97519898e-02 -8.77310666e-02 -8.66687826e-02 -8.15835644e-02 -7.32081259e-02 -6.02706951e-02 -4.24399368e-02 -2.10497270e-02 5.08304985e-04 2.06241750e-02 3.91292189e-02 5.77145097e-02 7.15135419e-02 7.45516578e-02 6.38238728e-02 4.35580179e-02 2.02712305e-02 -4.46692821e-03 -2.41242981e-02 -3.66799760e-02 -4.33587597e-02 -4.46705006e-02 -4.05827015e-02 -3.30819589e-02 -2.70683602e-02 -2.59245042e-02 -2.90609922e-02 -2.95609304e-02 -2.38858773e-02 -1.50880484e-02 -9.19969595e-03 -8.33417313e-03 -7.93651142e-03 -3.05461694e-03 7.13301998e-03 1.82026094e-02 2.66824286e-02 3.30622138e-02 4.00340894e-02 4.90608443e-02 6.02648286e-02 7.00942393e-02 7.31115967e-02 6.75011510e-02 5.35389324e-02 3.48700770e-02 1.69745209e-02 1.72454146e-03 -1.20611024e-02 -2.27968218e-02 -2.76882865e-02 -3.34147477e-02 -4.57197498e-02 -6.05078349e-02 -7.24893051e-02 -7.70420692e-02 -7.26272010e-02 -6.11005811e-02 -4.55265529e-02 -3.00707518e-02 -1.68398434e-02 -7.56863503e-03 -3.68954861e-03 -3.43061853e-03 -3.06004774e-03 2.55862832e-03 1.71655046e-02 4.20159830e-02 6.75693679e-02 8.23617564e-02 8.56835153e-02 8.13427954e-02 7.37373718e-02 6.28271473e-02 4.67098172e-02 2.60081154e-02 2.82900268e-03 -2.26376028e-02 -4.78755329e-02 -6.99617777e-02 -8.35343661e-02 -8.44844001e-02 -7.18198681e-02 -5.08207746e-02 -2.73111366e-02 -8.63745041e-03 2.79708853e-03 8.16043004e-03 7.92863690e-03 5.11229300e-03 3.49449875e-03 3.87115590e-03 2.85022886e-03 -1.51356430e-04 -2.71617609e-03 7.99093220e-04 1.28770189e-02 2.97511624e-02 4.23481503e-02 4.61655216e-02 4.49878002e-02 4.16959342e-02 3.51358326e-02 2.30338982e-02 7.71870254e-03 -6.49914172e-03 -1.95979278e-02 -2.91442689e-02 -3.34105565e-02 -3.22190201e-02 -2.83765449e-02 -2.52463396e-02 -2.32145504e-02 -2.10720714e-02 -1.84390441e-02 -1.47681365e-02 -1.01599056e-02 -4.95494275e-03 -3.16814647e-04 4.11336090e-03 7.87342088e-03 1.02960668e-02 1.29127543e-02 1.55124154e-02 1.78762099e-02 1.96778316e-02 2.12245457e-02 1.92318426e-02 1.52708347e-02 1.00532965e-02 3.93556366e-03 -2.00861226e-04 -3.92225374e-03 -5.30239979e-03 -4.82306995e-03 -4.47717083e-03 -6.19650760e-03 -1.00134814e-02 -1.42494639e-02 -1.54978616e-02 -1.35397676e-02 -1.13623819e-02 -9.41262041e-03 -7.58206760e-03 -5.06296458e-03 -2.48585060e-03 -4.00416183e-04 1.87050218e-03 3.62541073e-03 6.07766518e-03 7.21786016e-03 8.77502787e-03 9.95603332e-03 1.03558434e-02 9.73912155e-03 7.63280475e-03 5.35276100e-03 2.20653640e-03 -9.84994816e-05 -1.37884146e-03 -1.12248811e-03 -8.47538380e-04 -6.26076861e-04 -5.87386385e-04 -8.97700051e-04 -2.24337187e-03 -3.92268068e-03 -4.69266928e-03 -4.80954070e-03 -4.76843389e-03 -5.05903965e-03 -4.98518712e-03 -4.01197097e-03 -3.00891314e-03 -1.46903844e-03 3.53683094e-05 1.59446778e-03 3.12390382e-03 4.76890345e-03 5.59846498e-03 5.63295324e-03 5.09371953e-03 3.49773838e-03 1.79393075e-03 4.66522098e-04 -1.52569918e-04 -7.15600312e-04 -9.34996770e-04 -1.13022725e-03 -8.06959189e-04 -8.41610248e-04 -1.14800547e-03 -1.28965094e-03 -1.24884820e-03 -1.44400047e-03 -2.04579669e-03 -2.17836113e-03 -2.08363191e-03 -1.81376431e-03 -1.41088949e-03 -8.75421280e-04 -4.54586404e-04 8.29150513e-05 6.78654789e-04 1.04647304e-03 1.59651631e-03 1.75850262e-03 1.71743579e-03 1.72121567e-03 1.67674127e-03 1.26223831e-03 1.09938217e-03 6.79160864e-04 2.74810773e-04 2.51510300e-04 -1.81502999e-05 -2.49024384e-04 -6.74164538e-04 -7.71579948e-04 -8.51008782e-04 -7.46019036e-04 -7.91883682e-04 -1.02615516e-03 -1.27411149e-03 -1.14412599e-03 -1.10888173e-03 -1.24436555e-03 -5.70511174e-04 -2.67727213e-04 -5.10544602e-05 -7.44585105e-05 3.32434599e-04 6.16025534e-04 4.60911522e-04 4.52806510e-04 5.17149268e-04 8.86203824e-04 8.06249536e-04 7.62364216e-04 6.89168671e-04 6.48290053e-04 5.50507466e-04 5.00420100e-04 4.91031520e-04 1.90660816e-04 1.42033309e-04 7.11215401e-05 -4.54097265e-05 -3.46351385e-05 -3.10631849e-05 -9.24403309e-05 -1.47993095e-04 -2.10994331e-04 -2.82751339e-04 -2.75720314e-04 -2.00919076e-04 -1.47717375e-04 -6.51029766e-05 -3.22735306e-05 -1.15212367e-04 -1.80394605e-04 -1.91644236e-04 -2.09229079e-04 -2.87571537e-04 -2.91770309e-04 -3.03029185e-04 -3.65472126e-04 -2.40079196e-04 -2.57480565e-04 -2.13293922e-04 -2.23895882e-04 -2.05898811e-04 -2.51361361e-04 -2.15835935e-04 -3.26122661e-04 -3.60578368e-04 -3.48674339e-04 -2.51265522e-04 -9.42783246e-05 1.05407722e-05 1.35324496e-04 1.32768084e-04 2.38054795e-04 2.37051097e-04 2.60387811e-04 3.17263146e-04 4.05326800e-04 5.37145496e-04 6.98105861e-04 8.02577998e-04 8.44690107e-04 8.42747086e-04 9.18552333e-04]

It seems to work! For certainty, is this consistent with what you obtain?

Many thanks for your help, Matthias

________________________________________ Van: ifeffit-bounces@millenia.cars.aps.anl.gov [ifeffit-bounces@millenia.cars.aps.anl.gov] namens Matt Newville [newville@cars.uchicago.edu] Verzonden: woensdag 5 november 2014 14:23 Aan: XAFS Analysis using Ifeffit Onderwerp: Re: [Ifeffit] Larch-0.9.24 with Python2

Hi Matthias,

OK, having scipy 0.9.0 is almost certainly the problem. I'm not entirely sure how far back in scipy versions would be supported, but I believe that version 0.12 will work, and am certain that 0.13 will work. In fact, I'm pretty sure that version 0.13 is made several changes to the minimization package. I believe I may be relying on some of those changes for being able to use minimization method other than the default Levenberg-Marquardt least-squares algorithm (not that this is recommended for autobk() or feffit(), but it can sometimes be useful in other fitting problems).

The version of numpy you have should be not be a problem.

I'll set the required versions of scipy to 0.13 and numpy to 1.5.

On Wed, Nov 5, 2014 at 1:39 AM, Matthias Filez wrote:

...

Dear Matt, Bruce,

I verified the scipy version: 0.9.0. The installed numpy version is 1.6.1.

What versions are required?

Thanks a lot!

Matthias ________________________________________ Van: ifeffit-bounces@millenia.cars.aps.anl.gov [ifeffit-bounces@millenia.cars.aps.anl.gov] namens Matt Newville [newville@cars.uchicago.edu] Verzonden: dinsdag 4 november 2014 18:59 Aan: XAFS Analysis using Ifeffit Onderwerp: Re: [Ifeffit] Larch-0.9.24 with Python2

Hi Matthias,

Hmm, looking at this more closely, my guess is that you're using an older version scipy than I am. Can you verify what scipy version you have?

That is, I do see that the return value for this call did change sometime in the past 2 years, though I'm not sure exactly which version. I'll try to commit a patch to allow this older version to github today. But, I suspect that older scipy versions might cause other problems.

I should probably put in a more exact check for required scipy version.

--Matt _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

--Matt _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

--Matt