cmflow 0.1.1.dev57

Building TOUGH2/Waiwera models from layers of conceptual models

Install

pip install -U cmflow

Install Dependency

On Windows, Shapely and Rtree are easier to be installed by using Christoph Gohlke's non-official build:

• https://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely
• https://www.lfd.uci.edu/~gohlke/pythonlibs/#rtree

descartes can be installed on all platform by:

pip install descartes

On Linux (Ubuntu shown here) these can be installed via apt-get:

sudo apt-get install -y python-shapely
sudo apt-get install -y python-rtree
sudo apt-get install -y python-descartes

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.calc_bmstats(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)]

# find all blocks intersect with the zone
blocks, ratios = bm_geology.blocks_in_zone('BASE1')
block_idx, ratios = bm_geology.blocks_in_zone('BASE1', indices=True)

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, .

Base Model Stats, mainly numpy arrays with rows corresponding to mulgrid blocks, and columns corresponding to zones. Each is a value, usually between 0.0 and 1.0. Often 1.0 is indicating that particular block is fully within the zone.

.stats numpy array (n,m), n = num of model blocks, m = num of zones .zones list of zone names (str) .zonestats dict of stats column by zone names .cellstats dict of stats row by block name

6 elements, 3 zones
 A    B    C
0.0, 0.7, 0.3,  -> row sum to 1.0, element 0, 0.7 rock B, 0.3 rock C 
1.0, 0.0, 0.0, 
1.0, 0.0, 0.0, 
0.0, 0.5, 0.5, 
0.1, 0.2, 0.7, 
0.0, 1.0, 0.0, 
(this is only one way of using it, such as a rocktype)

.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