run_Cm_lookups.py
35e89619726amaster

run_Cm_lookups.py
View Options

			# -- coding: utf-8 --
			# @Author: Theo Lemaire
			# @Email: theo.lemaire@epfl.ch
			# @Date: 2020-08-05 17:42:12
			# @Last Modified by: Theo Lemaire
			# @Last Modified time: 2020-08-05 20:31:11

			''' Create Cm lookup table. '''

			import os
			import logging
			import numpy as np

			from PySONIC.utils import logger, isIterable, alert
			from PySONIC.core import BilayerSonophore, Batch, Lookup, AcousticDrive
			from PySONIC.parsers import MechSimParser


			@alert
			def computeCmLookup(bls, fref, Aref, mpi=False, loglevel=logging.INFO):
			descs = {'f': 'US frequencies', 'A': 'US amplitudes'}

			# Populate reference vectors dictionary
			refs = {
			'f': fref, # Hz
			'A': Aref # Pa
			}

			# Check validity of all reference vectors
			for key, values in refs.items():
			if not isIterable(values):
			raise TypeError(f'Invalid {descs[key]} (must be provided as list or numpy array)')
			if not all(isinstance(x, float) for x in values):
			raise TypeError(f'Invalid {descs[key]} (must all be float typed)')
			if len(values) == 0:
			raise ValueError(f'Empty {key} array')
			if key == 'f' and min(values) <= 0:
			raise ValueError(f'Invalid {descs[key]} (must all be strictly positive)')
			if key == 'A' and min(values) < 0:
			raise ValueError(f'Invalid {descs[key]} (must all be positive or null)')

			# Get references dimensions
			dims = np.array([x.size for x in refs.values()])

			# Create simulation queue
			drives = AcousticDrive.createQueue(fref, Aref)
			queue = [[drive, 0.] for drive in drives]

			# Run simulations and populate outputs
			logger.info(f'Starting Cm simulation batch for {bls}')
			batch = Batch(bls.getRelCmCycle, queue)
			rel_Cm_cycles = batch(mpi=mpi, loglevel=loglevel)

			# Make sure outputs size matches inputs dimensions product
			nout, nsamples = len(rel_Cm_cycles), rel_Cm_cycles[0].size
			assert nout == dims.prod(), 'Number of outputs does not match number of combinations'
			dims = np.hstack([dims, nsamples])
			refs['t'] = np.linspace(0., 1., nsamples)

			# Reshape effvars into nD arrays and add them to lookups dictionary
			logger.info('Reshaping output into lookup table')
			rel_Cm_cycles = np.array(rel_Cm_cycles).reshape(dims)

			# Construct and return lookup object
			return Lookup(refs, {'Cm_rel': rel_Cm_cycles})


			def main():

			parser = MechSimParser(outputdir='.')
			parser.addTest()
			parser.defaults['radius'] = 32.0 # nm
			parser.defaults['freq'] = np.array([20., 100., 500., 1e3, 2e3, 3e3, 4e3]) # kHz
			parser.defaults['amp'] = np.insert(
			np.logspace(np.log10(0.1), np.log10(600), num=50), 0, 0.0) # kPa
			args = parser.parse()
			logger.setLevel(args['loglevel'])

			# Model
			a = args['radius'][0] # m
			Cm0 = 1e-2 # F/m2
			Qm0 = 0.0 # C/m2
			bls = BilayerSonophore(a, Cm0, Qm0)
			lookup_fpath = bls.Cm_lkp_filepath

			# Batch inputs
			inputs = [args[x] for x in ['freq', 'amp']]

			# Adapt inputs and output filename if test case
			if args['test']:
			for i, x in enumerate(inputs):
			if x is not None and x.size > 1:
			inputs[i] = np.array([x.min(), x.max()])
			fcode, fext = os.path.splitext(lookup_fpath)
			lookup_fpath = f'{fcode}_test{fext}'

			# Check if lookup file already exists
			if os.path.isfile(lookup_fpath):
			logger.warning(
			f'"{lookup_fpath}" file already exists and will be overwritten. Continue? (y/n)')
			user_str = input()
			if user_str not in ['y', 'Y']:
			logger.error('Cm-lookup creation canceled')
			return

			# Compute lookup
			lkp = computeCmLookup(bls, *inputs, mpi=args['mpi'], loglevel=args['loglevel'])
			logger.info(f'Generated Cm-lookup: {lkp}')

			# Save lookup in PKL file
			logger.info(f'Saving {bls} Cm-lookup in file: "{lookup_fpath}"')
			lkp.toPickle(lookup_fpath)


			if __name__ == '__main__':
			main()

R4670/scripts/run_Cm_lookups.py35e89619726amaster

run_Cm_lookups.pyView Options

R4670/scripts/run_Cm_lookups.py
35e89619726amaster

run_Cm_lookups.py
View Options