[Hadstruct] Data locations

[Joseph Karpie]

Hey everyone,

I have locations of correlation functions from various times at places in ifarm and qcdi. I list many below for posterity.

Before that, I have been using the unix group lqcdpdf at qcdi with the hope to keep data only open to us. Could you please get yourself added to that group if you are not already? Run "groups" in command line to check. I am changing my files' groups to it at ifarm and I'd like to make it so the data is openly accessible to just us. I will be mucking with permissions, so let me know if you can't access something.

I would like to generally suggest as new data comes for the DA, gluon, and disco, that they go to a common place on ifarm and qcdi for safe keeping. I wish to continue to dump data into /work/JAM/jkarpie/data/ on ifarm. We have /cache/lqcdpdf/isoClover/ on qcdi which we used in the past for storing more sensitive things. I know Daniel puts some things there which is good.

Is it possible that someone with greater authority remove the other rwx permission from all the subdirectories of /cache/lqcdpdf/isoClover/? I changed mine and I can ask Daniel can change his (please do it) but that's all I can really control.

Best,
Joe



ifarm: At ifarm, I have placed all my analyses into a location under /work/JAM/jkarpie/. I suggest anyone intersted in using ifarm make such a work directory under /work/JAM/, which is actually a link to /w/jam-sciwork18/ which will appear in my pwd calls below. There can be easily shared amongst each other. In my work, you can find the fits for the evolution paper in /work/JAM/jkarpie/step_scaling_br/ or from JAM Latt in /work/JAM/jkarpie/jam_gluon/

Colin's distillation 2pt functions edbs and dat : In /w/jam-sciwork18/jkarpie/data/nuc_2pt/ there are subdirectories of the form momXYZ.#.#.# for the momentum of the 2pt functions.

Colin's unpolarized quark PDF matrix elements and correlators hdf5 : In /w/jam-sciwork18/jkarpie/data/pseudoPDF/unpol_quark_matelems/cl21_32_64_b6p3_m0p2350_m0p2050/ are hdf5 files with the correlators and fit results as well as one with the double ratio data.

Tanjib's gluon PDF  txt : /w/jam-sciwork18/jkarpie/data/pseudoPDF/gluon_corr/cl21_32_64_b6p3_m0p2350_m0p2050/nuc_unpol/data_op_mom_00#/ contains results of different fits for a given momentum extracted from the tarballs found in his qcdi /work. Switch unpol for pol to get helicity. They differ by combinations of index and number of operators used in the sGEVP. In them contain different gradient flow times. Ultimately the files which end in *sum_Rt+1_re.txt will contain the effective matrix element which plateus to the desired constant as T->inf.  For example, the polarized data with p=1, tau=1, z=2 for lorentz combination tx,yz and a fit with 3 ops is in
/w/jam-sciwork18/jkarpie/data/pseudoPDF/gluon_corr/cl21_32_64_b6p3_m0p2350_m0p2050/nuc_pol/data_op_mom_001/results_pPDF_pol_tx_yz_mom_1_op_3/results_wf_1.00/data_pPDF_pol_tx_yz_wf_1.0_mom_1_sep_2_start_1000_end_20000_mer_sum_Rt+1_re.txt

Colin's unpolarized quark GPD correlators edbs and dat : Within /w/jam-sciwork18/jkarpie/data/pseudoGPD/unpol_quark_matelems/cl21_32_64_b6p3_m0p2350_m0p2050/unphased  (and neigboring phased directory) are subdirectories labeled by sink and source momentum.  Within them are subdirectories for different source-sink separations as well as a fit_res which contains my fit results. In each fit_res directory are many files amplitude* which contain the A_i for that fit, including Herve's (I hope you doesn't mind. You now have rw access so run "rm /work/JAM/jkarpie/data/pseudoGPD/unpol_quark_matelems/cl21_32_64_b6p3_m0p2350_m0p2050/unphased/snk*src*/fit_res/amp*hd*dat" if you do.)


qcdi:

CLS Nf2 2pts and 3pts (pion/nuc PDFs and pion DAs) edbs : the subdirectories of /cache/lqcdpdf/isoClover/pseudo_data_storage/Nf2 label the ensemble. Follow subdirs to find edbs.

Good Lattice Cross Section mod : In directories such as /cache/lqcdpdf/isoClover/pionPDF-mod/b6p1/cl21_24_64_b6p1_m0p2800_m0p2450/pz01/pdf_m-0.2800_db/0/6000/ are the mod files which are the results of the GLCS runs. These contain the 2pt and 4 function functions for <pi(p) | O(x) O(x+z) |pi(p)> for ALL z with a randomly picked x. While costly compared to pseudo for on axis separations, having all z at fixed cost could be useful for a TMD like calculation.

Assorted others.
Colin has many correlators of various types including those above in /cache/lqcdpdf/isoClover/dist_*/ including the data for the things on ifarm.
Colin's unpolarized quark PDF correlators including the unpublished ensembles are in /cache/lqcdpdf/isoClover/dist_ppdf/
My PDF results from the old smearing are in /cache/lqcdpdf/isoClover/pPDF/ and I wish you luck sorting through my muck.
Tanjib's tarballs are also in /cache/lqcdpdf/isoClover/pPDF/gluon_correlators/tanjib/

--
Joe Karpie
Pronouns: He/His/Him
Postdoctoral Fellow
Theoretical and Computational Physics Center
Jefferson Lab

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/hadstruct/attachments/20231205/d40f72e1/attachment.html>