Investigate the PyNE material class to create a material library

[makeclean]

With @elliotbiondo investigate using PyNE to fully store, create and modify an engineering database of materials,

• [x] By reading in an MCNP xsdir file generate an internal availability of nuclides
• [x] Generate arbitrary material descriptions along with auxiliary data such as density, temperature etc
• [ ] Propose how we might produce a full description of each material for each transport code

[AliZeineddine]

Some comments about the below Code:

• command: python mat_desc -w 'xsdirpath' -o 'H5M file path' -p 'string of preferred libraries, white spaced'.
• '-o' option is not yet supported.
• '.h5' or '.text' file paths are temporarily hard coded until we extract the material names from the H5M file.
• While reading from '.txt' file, the isotopes in the composition will be outputted in the format "zzaaam" format. However it appears that the xsdir data is written in "zzaaa" format. Will adjust the code to the appropriate format once we decide which format will be used in the xsdir file in the future. Or maybe I will adjust the code to both formats.
• output: Dictionary of { 'isotope': 'library' } elements to be MCNP-written later.
• Some comments are in the code.

[AliZeineddine]

import linecache from optparse import OptionParser import os, sys from pyne.material import Material, from_atom_frac, from_hdf5, from_text, MapStrMaterial

def XSDIR_Search(file1,searchPhrase,preferredLibsStr): preferredLibs=preferredLibsStr.split() searchfile = open(file1, "r") searchPhraseDot=searchPhrase + '.' Lib=[]

for line in searchfile:
    if searchPhraseDot in line and line.find(searchPhraseDot)==1 or line.find(searchPhraseDot)==0:
    LineFound = line
    Lib.append( LineFound[LineFound.find(".")+1:LineFound.find(" ",LineFound.find(".")+1 )] )
print "Libraries for Isotope: ", searchPhrase, " found are: ", Lib
Iso_Lib=[]

if not Lib:
print " Isotope: ", searchPhrase, " has no Libraries found."
return 0
elif preferredLibsStr== "None":
Iso_Lib=[ searchPhrase, Lib[0]]
print "No preferred Libary, first to find was chosen: ", Iso_Lib
return Iso_Lib
else:
    for PLib in preferredLibs:
    if PLib in Lib:
        #Iso_Lib.append([ searchPhrase, PLib])
        Iso_Lib=[ searchPhrase, PLib]
        break
if not Iso_Lib:
    print "None of your preferred libraries was found for isotope: ", searchPhrase
    return 0
else:
       print "To MCNP: ", Iso_Lib
    return Iso_Lib

searchfile.close()

def main( arguments = None ):

#Instatiate options parser
parser = OptionParser\
    (usage='%prog <flux mesh> [options]')

parser.add_option('-w', dest='file1', default=None,\
    help='Path to the XSDIR file, default=%default')

parser.add_option('-o', dest='H5M_Path', default='None',\
    help='Path to the H5M file, default=%default')

parser.add_option('-p', dest='preferredLibsStr', default='None',\
    help='string of preffered libraries white space separated. Example: "21c 22c 67c", default=%default')
OPT, ARG = parser.parse_args()

Dictionary={}
Mat_Comp_List=[]
##### Mainly we want for-loop through the different '.h5' or '.text' files corresponding to each material name we get from the '.h5m' file
##### to have more elements in the Mat_Comp_List
####Example: Mat_Comp_list= [ [Mat1, iso1, iso2, iso3], [Mat2 , iso1, iso2 ,iso3], ...]

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# Reading '.h5' file to extract isotopes of the material
#mat = Material()
#mat.from_hdf5("/home/zeineddine/PYNE/Pyne/pyne-pyne-e14f89d/pyne/tests/mat.h5", "/mat", 1, protocol=0)
#-------------------------------------------------------------------------------------------------------------------------

#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##### Or directly Read a text file of format: example: #####
##### Name    SomeMaterial
##### Mass    42
##### APerM   1
##### H1    0.04
##### U238    0.9
##### 6012    0.06
mat = Material("/home/zeineddine/PYNE/Pyne/pyne-pyne-e14f89d/pyne/tests/SomeMaterial.txt")
#-------------------------------------------------------------------------------------------------------------------------

print "Composition of tempate '.h5' or '.text' file: ", mat.comp
Isotopes=[]
for key in mat.comp.keys():
Isotopes.append(str(key))
Isotopes.insert(0,'materialName')

Mat_Comp_List.append(Isotopes)
##### Mat_Comp_list now will look like := [ [Mat1, iso1, iso2, iso3] ]
print "list of 'Material,Composition' lists: ", Mat_Comp_List

for list in Mat_Comp_List:
for iso in list:
    if iso !=list[0]:
        ResultDictElement=XSDIR_Search(OPT.file1,iso,OPT.preferredLibsStr) 
        ##### XSDIR_Search: RETURN example: ['6012', '90y'] if found, 0 if not found
        if ResultDictElement !=0:
            Dictionary.update({ResultDictElement[0]:ResultDictElement[1]})
print "Dictionary: ", Dictionary

if name == 'main': main()

[gonuke]

It may be more valuable to check this code into the repository. Then others can check it out and play with it and make comments there.

[AliZeineddine]

I uploaded another file 'mat_desc2.py' . It was added in a branch called patch-2.

We are still working on extracting the material names from the '.h5m' file. In the meantime, I am temporarily using the 'retag_hdf5' function through python. That is I am outputting the material tags to the command line and then getting them back and using them in the Python Script. This is just for the sake of completing the other parts of the script. In the new Script:

• I assumed that the last unused command line output is always "OBB 2", which happened to be the case in all the h5m files I ran through 'retag_hdf5'.
• the '-o' option is now supported. {h5m Path}
• The code will be searching for the material library files ( '.h5' or '.txt' ) located in a certain hard-coded folder path, In my case :"/home/zeineddine/Mat_Libs/"
• The 'retag_hdf5.cpp' file path is also hard-coded

[makeclean]

@AliZeineddine I have just pushed a pytaps script to replace the retag_hdf5 hack. Look for commit edced054fcec305dab899492d9eb48ffe6ebab15, you will need to pull from this repo. Please also add your other script to the repo.

[makeclean]

Closed will pull c9495f956ddda0a5cf53affd9d14c3780066fc18