cmflow 0.1.1.dev8

Building TOUGH2/Waiwera models from layers of conceptual models

# Example

Creates BMStats that can be used later, from Leapfrog Geology:

# (ONLY ONCE) geo used to get geology from Leapfrog geological model cmgeo = mulgrid(‘g_very_fine.dat’)

# CSV file created by Leapfrog using cmgeo above leapfrog = LeapfrogGM() leapfrog.import_leapfrog_csv(‘grid_gtmp_ay2017_03_6_fit.csv’)

cm_geology = CM_Blocky(cmgeo, leapfrog)

# whatever active model we are working on bmgeo = mulgrid(‘gwaixx_yy.dat’)

bms_geology = cm_geology.populate_model(bm_geo) bms_geology.save(‘a.json’)

A BMStats object can be reused (very fast) to eg.

bms_geology = BMStats(‘a.json’)

# get a cell’s stats cs = bms_geology.cellstats[‘abc12’]

# rock that occupies most in cell ‘abc12’ rock_name = bms_geology.zones[np.argmax(cs)]

# how many rock in cell ‘abc12’ n_rock = len(np.nonzero(cs))

# list all rocks in cell ‘abc12’ rocks = [bm_geology.zones[i] for i in np.nonzero(cs)]

# BMStats

This is the object that we keep for later use. It is associated to a certain “geometry” file. So each cell has information on zones. Usually this is generated by cm.populate_model(), which can be expensive.

• ? should I call it CMStats?

• ? TODO, .cellstats access by cell index

• ? TODO, .

.stats, numpy array (n * m), n number of geometry cells, m number of zones .zones, a list of zone name, eg. geology rock names, fault names etc .zonestats, a dict keyed by zone name, an array of size number of cells, each cell is between .cellstats, a dict of stats by cell name

.save() .load() .add_stats() add another bmstat, merge stats .add_cm() calls cm.populate_model, and merge stats

# CM # CM_Blocky # CM_Prism # CM_Faults

These are the objects that can be created in order to create the final BMStats objects. The common method .populate_model(bm_geo) is called to create BMStats objects. It means the conceptual model is “applied” onto the bm_geo.

• TODO, .populate_model() should return BMStats instead

• ? TODO, .populate_model() should be called something else?

.populate_model(bm_geo) takes a target geometry, and return/creates BMStats

# LeapfrogGM