def read_xyz_n_dm(fxyz, design_matrix, use_atomic_descriptors, only_use_species, peratom):
dm = []
dm_atomic = []
# try to read the xyz file
if fxyz is not None and fxyz != 'none':
from asaplib.data import ASAPXYZ
# from asapxyzs import ASAPXYZs
asapxyz = ASAPXYZ(fxyz)
if use_atomic_descriptors:
dm = asapxyz.get_atomic_descriptors(design_matrix, only_use_species)
else:
dm, dm_atomic = asapxyz.get_descriptors(design_matrix, peratom)
else:
asapxyz = None
print("Did not provide the xyz file. We can only output descriptor matrix.")
# we can also load the descriptor matrix from a standalone file
import os
if os.path.isfile(design_matrix[0]):
try:
import numpy as np
dm = np.genfromtxt(design_matrix[0], dtype=float)
print("loaded the descriptor matrix from file: ", design_matrix[0])
except:
raise ValueError('Cannot load the descriptor matrix from file')
return asapxyz, dm, dm_atomic
def main():
"""
Test if Ridge regression is working.
Parameters
----------
fxyz: string giving location of xyz file
prefix: string giving the filename prefix
"""
fxyz = os.path.join(os.path.split(__file__)[0], 'small_molecules-SOAP.xyz')
fmat = ['SOAP-n4-l3-c1.9-g0.23']
fy = 'dft_formation_energy_per_atom_in_eV'
prefix = "test-skrr"
test_ratio = 0.05
lc_points = 8
lc_repeats = 8
# try to read the xyz file
asapxyz = ASAPXYZ(fxyz)
desc, _ = asapxyz.get_descriptors(fmat, False)
y_all = asapxyz.get_property(fy)
# print(desc)
dm = Design_Matrix(X=desc, y=y_all, whiten=True, test_ratio=test_ratio)
# kernel, jitter, delta, sigma, sparse_mode="fps", n_sparse=None
k_spec = {
'k0': {
"type": "linear"
}
} # { 'k1': {"type": "polynomial", "d": power}}
# if sigma is not set...
sigma = 0.001 * np.std(y_all)
krr = KRRSparse(0., None, sigma)
skrr = SPARSE_KRR_Wrapper(k_spec, krr, sparse_mode="fps", n_sparse=-1)
# fit the model
dm.compute_fit(skrr, 'skrr', store_results=True, plot=True)
# learning curve
if lc_points > 1:
dm.compute_learning_curve(skrr,
'ridge_regression',
lc_points=lc_points,
lc_repeats=lc_repeats,
randomseed=42,
verbose=False)
dm.save_state(prefix)
plt.show()
def main():
"""
Test if Ridge regression is working.
Parameters
----------
fxyz: string giving location of xyz file
prefix: string giving the filename prefix
"""
fxyz = os.path.join(os.path.split(__file__)[0], 'small_molecules-SOAP.xyz')
fmat = ['SOAP-n4-l3-c1.9-g0.23']
fy = 'dft_formation_energy_per_atom_in_eV'
prefix = "test-rr"
test_ratio = 0.05
lc_points = 8
lc_repeats = 8
# try to read the xyz file
asapxyz = ASAPXYZ(fxyz)
desc, _ = asapxyz.get_descriptors(fmat, False)
y_all = asapxyz.get_property(fy)
# print(desc)
dm = Design_Matrix(X=desc, y=y_all, whiten=True, test_ratio=test_ratio)
# if sigma is not set...
sigma = 0.001 * np.std(y_all)
rr = RidgeRegression(sigma)
# fit the model
dm.compute_fit(rr, 'ridge_regression', store_results=True, plot=True)
# learning curve
if lc_points > 1:
dm.compute_learning_curve(rr,
'ridge_regression',
lc_points=lc_points,
lc_repeats=lc_repeats,
randomseed=42,
verbose=False)
dm.save_state(prefix)
plt.show()
def main():
"""
Select frames from the supplied xyz file (fxyz) using one of the following algorithms:
1. random: random selection
2. fps: farthest point sampling selection. Need to supply a kernel matrix or descriptor matrix using -fmat
4. CUR decomposition
Parameters
----------
fxyz: Path to xyz file.
fmat: Path to the design matrix or name of the tags in ase xyz file
prefix: Filename prefix, default is ASAP
nkeep: The number of representative samples to select
algorithm: 'the algorithm for selecting frames ([random], [fps], [cur])')
fmat: Location of descriptor or kernel matrix file. Needed if you select [fps] or [cur].
"""
fxyz = os.path.join(os.path.split(__file__)[0], 'small_molecules-SOAP.xyz')
fmat = ['SOAP-n4-l3-c1.9-g0.23']
nkeep = 10
prefix = "test-frame-select"
# read the xyz file
asapxyz = ASAPXYZ(fxyz)
# for both algo we read in the descriptor matrix
desc, _ = asapxyz.get_descriptors(fmat)
print("shape of the descriptor matrix: ", np.shape(desc),
"number of descriptors: ", np.shape(desc[0]))
for algorithm in ['random', 'cur', 'fps']:
sparsifier = Sparsifier(algorithm)
sbs = sparsifier.sparsify(desc, nkeep)
# save
selection = np.zeros(asapxyz.get_num_frames(), dtype=int)
for i in sbs:
selection[i] = 1
np.savetxt(prefix + "-" + algorithm + "-n-" + str(nkeep) + '.index',
selection,
fmt='%d')
asapxyz.write(prefix + "-" + algorithm + "-n-" + str(nkeep), sbs)
def main(fmat, fxyz, fy, prefix, scale, test_ratio, sigma, lc_points,
lc_repeats):
"""
Parameters
----------
fmat: Location of descriptor matrix file or name of the tags in ase xyz file. You can use gen_descriptors.py to compute it.
fxyz: Location of xyz file for reading the properties.
fy: Location of property list (1D-array of floats)
prefix: filename prefix for learning curve figure
scale: Scale the coordinates (True/False). Scaling highly recommanded.
test_ratio: train/test ratio
sigma: noise level in kernel ridge regression, default is 0.1% of the standard deviation of the data.
lc_points : number of points on the learning curve
lc_repeats : number of sub-sampling when compute the learning curve
Returns
-------
Learning curve.
"""
scale = bool(scale)
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, _ = asapxyz.get_descriptors(fmat)
# we can also load the descriptor matrix from a standalone file
if os.path.isfile(fmat[0]):
try:
desc = np.genfromtxt(fmat[0], dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
if len(desc) == 0:
raise ValueError(
'Please supply descriptor in a xyz file or a standlone descriptor matrix'
)
print("shape of the descriptor matrix: ", np.shape(desc),
"number of descriptors: ", np.shape(desc[0]))
# read in the properties to be predicted
y_all = []
try:
y_all = np.genfromtxt(fy, dtype=float)
except:
y_all = asapxyz.get_property(fy)
dm = Design_Matrix(X=desc, y=y_all, whiten=True, test_ratio=test_ratio)
# if sigma is not set...
if sigma < 0:
sigma = 0.001 * np.std(y_all)
rr = RidgeRegression(sigma)
# fit the model
dm.compute_fit(rr, 'ridge_regression', store_results=True, plot=True)
# learning curve
if lc_points > 1:
lc_scores = dm.compute_learning_curve(rr,
'ridge_regression',
lc_points=lc_points,
lc_repeats=lc_repeats,
randomseed=42,
verbose=False)
# make plot
lc_scores.plot_learning_curve()
plt.show()
def main(fmat, fxyz, ftags, prefix, dimension, pc1, pc2, adtext):
"""
Parameters
----------
fmat: Location of low-dimensional coordinate file.
ftags: Location of tags for the first M samples.
prefix: Filename prefix.
dimension: Number of the first X dimensions to keep
pc1: First principle axis (int)
pc2: Second principle axis (int)
adtext: Boolean giving whether to adjust text or not.
Returns
-------
"""
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, _ = asapxyz.get_descriptors(fmat)
if os.path.isfile(fmat):
try:
desc = np.genfromtxt(fmat, dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
if len(desc) == 0:
raise ValueError('Please supply descriptor in a xyz file or a standlone descriptor matrix')
print("loaded", fmat, " with shape", np.shape(desc))
# load tags if any
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")
ndict = len(tags)
proj = np.asmatrix(desc)[:, 0:dimension]
density_model = KDE_internal() # KDE_sklearn(bandwidth=1) # KDE_scipy()
# fit density model to data
try:
density_model.fit(proj)
except:
raise RuntimeError('KDE did not work. Try smaller dimension.')
rho = density_model.evaluate_density(proj)
# save the density
np.savetxt(prefix + "-kde.dat", np.transpose([np.arange(len(rho)), rho]),
header='index log_of_kernel_density_estimation', fmt='%d %4.8f')
# color scheme
plotcolor = rho
colorlabel = 'Log of density for every point'
[plotcolormin, plotcolormax] = [np.min(plotcolor), np.max(plotcolor)]
# make plot
plot_styles.set_nice_font()
# density plot
fig, ax = plot_styles.plot_density_map(np.asarray(proj[:, [pc1, pc2]]), plotcolor,
xlabel='Princple Axis ' + str(pc1), ylabel='Princple Axis ' + str(pc2),
clabel=colorlabel, label=None,
xaxis=True, yaxis=True,
centers=None,
psize=None,
out_file=None,
title='KDE for: ' + prefix,
show=False, cmap='gnuplot',
remove_tick=False,
use_perc=False,
rasterized=True,
fontsize=15,
vmax=plotcolormax,
vmin=plotcolormin)
fig.set_size_inches(18.5, 10.5)
if ftags != 'none':
texts = []
for i in range(ndict):
if tags[i] != 'None' and tags[i] != 'none' and tags[i] != '':
ax.scatter(proj[i, pc1], proj[i, pc2], marker='^', c='black')
texts.append(ax.text(proj[i, pc1], proj[i, pc2], tags[i],
ha='center', va='center', fontsize=15, color='red'))
# ax.annotate(tags[i], (proj[i,pc1], proj[i,pc2]))
if adtext:
from adjustText import adjust_text
adjust_text(texts, on_basemap=True, # only_move={'points':'', 'text':'x'},
expand_text=(1.01, 1.05), expand_points=(1.01, 1.05),
force_text=(0.03, 0.5), force_points=(0.01, 0.25),
ax=ax, precision=0.01,
arrowprops=dict(arrowstyle="-", color='black', lw=1, alpha=0.8))
plt.show()
fig.savefig('kde_4_' + prefix + '.png')
def main(fmat, fxyz, ftags, fcolor, colorscol, prefix, output, kpca_d, pc1,
pc2, adtext):
"""
Parameters
----------
fmat
fxyz
ftags
fcolor
colorscol
prefix
output
kpca_d: number of dimensions
pc1
pc2
adtext
Returns
-------
"""
foutput = prefix + "-kpca-d" + str(kpca_d)
# load the kernel matrix
try:
kNN = np.genfromtxt(fmat, dtype=float)
except:
raise ValueError('Cannot load the kernel matrix')
print("loaded", fmat)
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")
if tags.ndim > 1:
tags = tags[:, 0]
ndict = len(tags)
asapxyz = None
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
elif output == 'xyz':
print(
"Did not provide the xyz file. We can only output descriptor matrix."
)
output = 'matrix'
# main thing
proj = KernelPCA(kpca_d).fit_transform(kNN)
# save
if output == 'matrix':
np.savetxt(prefix + "-kpca-d" + str(kpca_d) + ".coord",
proj,
fmt='%4.8f',
header='low D coordinates of samples')
elif output == 'xyz':
if os.path.isfile(foutput + ".xyz"):
os.rename(foutput + ".xyz", "bck." + foutput + ".xyz")
asapxyz.set_descriptors(proj, 'kpca_coord')
asapxyz.write(foutput)
# color scheme
plotcolor, colorlabel, colorscale = set_color_function(
fcolor, asapxyz, colorscol, len(proj))
# make plot
plot_styles.set_nice_font()
# fig, ax = plt.subplots()
fig, ax = plot_styles.plot_density_map(proj[:, [pc1, pc2]],
plotcolor,
xlabel='Principal Axis ' + str(pc1),
ylabel='Principal Axis ' + str(pc2),
clabel=colorlabel,
label=None,
xaxis=True,
yaxis=True,
centers=None,
psize=None,
out_file=None,
title='KPCA for: ' + prefix,
show=False,
cmap='gnuplot',
remove_tick=False,
use_perc=True,
rasterized=True,
fontsize=15,
vmax=colorscale[1],
vmin=colorscale[0])
fig.set_size_inches(18.5, 10.5)
if ftags != 'none':
texts = []
for i in range(ndict):
if tags[i] != 'None' and tags[i] != 'none' and tags[i] != '':
ax.scatter(proj[i, pc1], proj[i, pc2], marker='^', c='black')
texts.append(
ax.text(proj[i, pc1],
proj[i, pc2],
tags[i],
ha='center',
va='center',
fontsize=15,
color='red'))
# ax.annotate(tags[i], (proj[i,pc1], proj[i,pc2]))
if adtext:
from adjustText import adjust_text
adjust_text(
texts,
on_basemap=True, # only_move={'points':'', 'text':'x'},
expand_text=(1.01, 1.05),
expand_points=(1.01, 1.05),
force_text=(0.03, 0.5),
force_points=(0.01, 0.25),
ax=ax,
precision=0.01,
arrowprops=dict(arrowstyle="-", color='black', lw=1,
alpha=0.8))
plt.show()
fig.savefig('KPCA_4_' + prefix + '-c-' + fcolor + '.png')
parser.add_argument("--stride","-s",type=int,help="stride for asap gen_desc command; this function will be deprecated")
parser.add_argument("--method","-m",type=str,default='fps',help="method, 3 options: 'random', 'cur', 'fps'")
args = parser.parse_args()
#dirctory = '/Users/jiedeng/Documents/tmp/jd848/project_folder/liquid_vapor/water1/r6-6k/cont1/asap'
#fxyz = dirctory+'/ASAP-desc.xyz'
fxyz = args.input
#fxyz = os.path.join(os.path.split(__file__)[0], 'small_molecules-SOAP.xyz')
# fmat = ['SOAP-n4-l3-c1.9-g0.23']
fmat = ['*']
nkeep = args.number #50
prefix = "test-frame-select"
# read the xyz file
asapxyz = ASAPXYZ(fxyz)
# for both algo we read in the descriptor matrix
desc, _ = asapxyz.get_descriptors(fmat)
print("shape of the descriptor matrix: ", np.shape(desc), "number of descriptors: ", np.shape(desc[0]))
algorithm = args.method#'fps' # 3 options: 'random', 'cur', 'fps'
#algorithm = 'random' # 3 options: 'random', 'cur', 'fps'
sparsifier = Sparsifier(algorithm)
sbs = sparsifier.sparsify(desc, nkeep)
sbs.sort()
if args.stride is None:
pass
else:
sbs = sbs*args.stride
def main(fxyz, dictxyz, prefix, output, peratom, fsoap_param, soap_rcut, soap_g, soap_n, soap_l, zeta_list, kernel_type, element_wise, soap_periodic, stride):
"""
Generate the SOAP descriptors.
Parameters
----------
fxyz: string giving location of xyz file
prefix: string giving the filename prefix
output: [xyz]: append the SOAP descriptors to extended xyz file; [mat] output as a standlone matrix
fsoap_param: use (possibly multiple sets) of SOAP descriptors using parameters specified in fsoap_param file (json format)
soap_rcut: float giving the cutoff radius, default value is 3.0
soap_g: float giving the atom width
soap_n: int giving the maximum radial label
soap_l: int giving the maximum angular label. Must be less than or equal to 9
zeta_list : get the global descriptor from atomic ones of zeta th power
kernel_type: type of operations to get global descriptors from the atomic soap vectors
elementwise: consider different species seperately when computing global descriptors from the atomic soap vectors
soap_periodic: string (True or False) indicating whether the system is periodic
stride: compute descriptor each X frames
"""
# read frames
asapxyz = ASAPXYZ(fxyz)
if fsoap_param is not None:
import json
# load the parameter from json file
if os.path.isfile(fsoap_param):
try:
with open(fsoap_param, 'r') as soapfile:
soap_js = json.load(soapfile)
except:
raise IOError('Cannot load the json file for soap parameters')
# use the default parameters
else:
soap_js = universal_soap_hyper(global_species, fsoap_param, dump=True)
# make descriptors
soap_desc_atomic = []
for element in soap_js.keys():
soap_param = soap_js[element]
[species_now, cutoff_now, g_now, n_now, l_now] = [soap_param['species'], soap_param['cutoff'],
soap_param['atom_gaussian_width'], soap_param['n'],
soap_param['l']]
soap_desc_atomic.append(SOAP(species=species_now, rcut=cutoff_now, nmax=n_now, lmax=l_now,
sigma=g_now, rbf="gto", crossover=False, average=False,
periodic=soap_periodic))
foutput = prefix + "-soapparam" + '-' + fsoap_param
desc_name = "SOAPPARAM" + '-' + fsoap_param
else:
soap_desc_atomic = [SOAP(species=global_species, rcut=soap_rcut, nmax=soap_n, lmax=soap_l,
sigma=soap_g, rbf="gto", crossover=False, average=False, periodic=soap_periodic)]
foutput = prefix + "-n" + str(soap_n) + "-l" + str(soap_l) + "-c" + str(soap_rcut) + "-g" + str(soap_g)
desc_name = "SOAP" + "-n" + str(soap_n) + "-l" + str(soap_l) + "-c" + str(soap_rcut) + "-g" + str(soap_g)
for i, frame in enumerate(frames):
fnow = soap_desc_atomic[0].create(frame, n_jobs=8)
for soap_desc_atomic_now in soap_desc_atomic[1:]:
fnow = np.append(fnow, soap_desc_atomic_now.create(frame, n_jobs=8), axis=1)
if kernel_type == 'average' and element_wise == False and len(zeta_list)==1 and zeta_list[0]==1:
# this is the vanilla situation. We just take the average soap for all atoms
frame.info[desc_name] = Atomic_2_Global_Descriptor_By_Species(fnow, [], [], kernel_type, zeta_list)
elif element_wise == False:
frame.info[desc_name+'-'+kernel_type] = Atomic_2_Global_Descriptor_By_Species(fnow, [], [], kernel_type, zeta_list)
else:
frame.info[desc_name+'-'+kernel_type+'-elementwise'] = Atomic_2_Global_Descriptor_By_Species(fnow, frame.get_atomic_numbers(), global_species, kernel_type, zeta_list)
# save
if output == 'matrix':
asapxyz.write_descriptor_matrix(desc_name, desc_name)
if peratom or nframes == 1:
asapxyz.write_atomic_descriptor_matrix(desc_name, desc_name)
elif output == 'xyz':
asapxyz.write(foutput)
else:
raise ValueError('Cannot find the output format')
def load_asapxyz(data_spec):
from asaplib.data import ASAPXYZ
return ASAPXYZ(data_spec['fxyz'], data_spec['stride'],
data_spec['periodic'], data_spec['fxyz_format'])
def main(fmat, fxyz, fy, prefix, test_ratio, jitter, n_sparse, sigma,
lc_points, lc_repeats):
"""
Parameters
----------
fmat: Location of kernel matrix file.
fy: Location of property list (1D-array of floats)
prefix: filename prefix for learning curve figure
test_ratio: train/test ratio
jitter: jitter level, default is 1e-10
n_sparse: number of representative samples, default is 5% of the data
sigma: noise level in kernel ridge regression, default is 0.1% of the standard deviation of the data.
lc_points : number of points on the learning curve
lc_repeats : number of sub-sampling when compute the learning curve
Returns
-------
Fitting outcome & Learning curve.
"""
# if it has been computed before we can simply load it
try:
K_all = np.genfromtxt(fmat, dtype=float)
except OSError:
raise Exception(
'fmat file could not be loaded. Please check the filename')
print("loaded", fmat)
# read in the properties to be predicted
y_all = []
try:
y_all = np.genfromtxt(fy, dtype=float)
except:
try:
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
y_all = asapxyz.get_property(fy)
except OSError:
raise Exception(
'property vector file could not be loaded. Please check the filename'
)
if len(y_all) != len(K_all):
raise ValueError(
'Length of the vector of properties is not the same as number of samples'
)
else:
n_sample = len(K_all)
# train test split
if test_ratio > 0:
K_train, K_test, y_train, y_test, _, _ = kernel_random_split(
K_all, y_all, test_ratio)
else:
K_train = K_test = K_all
y_train = y_test = y_all
n_train = len(K_train)
n_test = len(K_test)
# set default value of n_sparse
if n_sparse == 0:
n_sparse = n_train // 20
# sparsification
if n_sparse >= n_train:
print(
"the number of representative structure is too large, please select n < ",
n_train)
elif n_sparse > 0:
ifps, dfps = fps(K_train, n_sparse, 0)
K_MM = K_train[:, ifps][ifps]
K_NM = K_train[:, ifps]
K_TM = K_test[:, ifps]
else:
print("it's usually better to use some sparsification")
K_MM = K_train
K_NM = K_train
K_TM = K_test
# if sigma is not set...
if sigma < 0:
sigma = 0.001 * np.std(y_train)
delta = np.std(y_train) / (np.trace(K_MM) / len(K_MM))
krr = KRRSparse(jitter, delta, sigma)
# fit the model
krr.fit(K_MM, K_NM, y_train)
fit_error = {}
# get the predictions for train set
y_pred = krr.predict(K_NM)
# compute the CV score for the dataset
y_pred, y_pred_test, fit_error = krr.get_train_test_error(K_NM,
y_train,
K_TM,
y_test,
verbose=True,
return_pred=True)
# dump to file
import json
with open('KRR_train_test_errors_4' + prefix + '.json', 'w') as fp:
json.dump(fit_error, fp)
# learning curve
# decide train sizes
if lc_points > 1 and n_sparse > 0:
train_sizes = exponential_split(n_sparse, n_train - n_test, lc_points)
print("Learning curves using train sizes: ", train_sizes)
lc_stats = lc_repeats * np.ones(lc_points, dtype=int)
lc = LCSplit(ShuffleSplit,
n_repeats=lc_stats,
train_sizes=train_sizes,
test_size=n_test,
random_state=10)
lc_scores = LC_SCOREBOARD(train_sizes)
for lctrain, _ in lc.split(y_train):
Ntrain = len(lctrain)
lc_K_NM = K_NM[lctrain, :]
lc_y_train = y_train[lctrain]
# here we always use the same test set
# otherwise, one can do `lc_K_test = K_NM[lctest,:]; lc_y_test = y_train[lctest]`
krr.fit(K_MM, lc_K_NM, lc_y_train)
# here we always use the same test set
_, lc_score_now = krr.fit_predict_error(K_MM, lc_K_NM, lc_y_train,
K_TM, y_test)
lc_scores.add_score(Ntrain, lc_score_now)
sc_name = 'RMSE' # MAE, RMSE, SUP, R2, CORR
lc_results = lc_scores.fetch(sc_name)
# output learning curve
np.savetxt("KRR_learning_curve_4" + prefix + ".dat", lc_results)
plot_styles.set_nice_font()
if lc_points > 1 and n_sparse > 0:
fig = plt.figure(figsize=(8 * 2.1, 8))
ax = fig.add_subplot(121)
else:
fig = plt.figure(figsize=(8, 8))
ax = fig.add_subplot(111)
ax.plot(y_train, y_pred, 'b.', label='train')
ax.plot(y_test, y_pred_test, 'r.', label='test')
ax.legend()
ax.set_title('KRR for: ' + fy)
ax.set_xlabel('actual y')
ax.set_ylabel('predicted y')
if lc_points > 1 and n_sparse > 0:
ax2 = fig.add_subplot(122)
ax2.errorbar(lc_results[:, 0],
lc_results[:, 1],
yerr=lc_results[:, 2],
linestyle='',
uplims=True,
lolims=True)
ax2.set_title('Learning curve')
ax2.set_xlabel('Number of training samples')
ax2.set_ylabel('Test {}'.format(sc_name))
ax2.set_xscale('log')
ax2.set_yscale('log')
plt.show()
fig.savefig('KRR_4_' + prefix + '.png')
r1_tag = np.ones(100).astype(int)
r3_tag = (np.ones(250) * 0).astype(int)
ppv_tag = (np.ones(75) * 2).astype(int)
r1_r3_tag = np.concatenate((r1_tag, r3_tag))
r1_r3_ppv_tag = np.concatenate((r1_tag, r3_tag, ppv_tag))
np.savetxt('tag', r1_r3_tag, fmt='%d')
np.savetxt('tag_ppv', r1_r3_ppv_tag, fmt='%d')
#fmat = 'pca_coord'
fmat = 'skpca-d-10'
#fmat = '[*]'
asapxyz = ASAPXYZ(fxyz)
dm, _ = asapxyz.get_descriptors(fmat, False)
dm_mg = asapxyz.get_atomic_descriptors(fmat, 12)
dm_oxygen = asapxyz.get_atomic_descriptors(fmat, 8)
dm_silicon = asapxyz.get_atomic_descriptors(fmat, 14)
plotcolor_volume, _, _, _ = set_color_function('volume', asapxyz)
plotcolor_density = np.zeros(len(plotcolor_volume))
for i in range(len(plotcolor_volume)):
plotcolor_density[i] = 29.889703 / plotcolor_volume[i] / 3.
#tags = np.loadtxt('ice-54-labels.dat', dtype="str")[:,0]
#iceornot_hydrogen, _, _, _ = set_color_function('ice-or-not.tag', asapxyz, 0, 0, False, True, 1, False)
iceornot_oxygen, _, _, _ = set_color_function('tag', asapxyz, 0, 0, False,
def main(fmat, fxyz, ftags, fcolor, colorscol, prefix, output, peratom, keepraw, scale, pca_d, pc1, pc2, projectatomic, plotatomic,
adtext):
"""
Parameters
----------
fmat: Location of descriptor matrix file or name of the tags in ase xyz file. You can use gen_descriptors.py to compute it.
fxyz: Location of xyz file for reading the properties.
ftags: Location of tags for the first M samples. Plot the tags on the PCA map.
fcolor: Location of a file or name of the tags in ase xyz file. It should contain properties for all samples (N floats) used to color the scatterplot'
colorscol: The column number of the properties used for the coloring. Starts from 0.
prefix: Filename prefix, default is ASAP
output: The format for output files ([xyz], [matrix]). Default is xyz.
peratom: Whether to output per atom pca coordinates (True/False)
keepraw: Whether to keep the high dimensional descriptor when output is an xyz file (True/False)
scale: Scale the coordinates (True/False). Scaling highly recommanded.
pca_d: Number of the principle components to keep
pc1: Plot the projection along which principle axes
pc2: Plot the projection along which principle axes
projectatomic: build the projection using the (big) atomic descriptor matrix
plotatomic: Plot the PCA coordinates of all atomic environments (True/False)
adtext: Whether to adjust the texts (True/False)
Returns
-------
"""
foutput = prefix + "-pca-d" + str(pca_d)
use_atomic_desc = (peratom or plotatomic or projectatomic)
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, desc_atomic = asapxyz.get_descriptors(fmat, use_atomic_desc)
if projectatomic: desc = desc_atomic.copy()
else:
asapxyz = None
print("Did not provide the xyz file. We can only output descriptor matrix.")
output = 'matrix'
# we can also load the descriptor matrix from a standalone file
if os.path.isfile(fmat[0]):
try:
desc = np.genfromtxt(fmat[0], dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
# sanity check
if len(desc) == 0:
raise ValueError('Please supply descriptor in a xyz file or a standlone descriptor matrix')
print("shape of the descriptor matrix: ", np.shape(desc), "number of descriptors: ", np.shape(desc[0]))
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")[:]
ndict = len(tags)
else:
tags = []
reduce_dict = { "pca":
{"type": 'PCA', 'parameter':{"n_components": pca_d, "scalecenter": scale}}
}
"""
reduce_dict = { "umap":
{"type": 'UMAP', 'parameter':{"n_components": pca_d, "n_neighbors": 10}}
}
reduce_dict = {
"reduce1_pca": {"type": 'PCA', 'parameter':{"n_components": 20, "scalecenter":True}},
"reduce2_tsne": {"type": 'TSNE', 'parameter': {"n_components": 2, "perplexity":20}}
}
"""
dreducer = Dimension_Reducers(reduce_dict)
proj = dreducer.fit_transform(desc)
if peratom or plotatomic and not projectatomic:
proj_atomic_all = dreducer.transform(desc_atomic)
# save
if output == 'matrix':
np.savetxt(foutput + ".coord", proj, fmt='%4.8f', header='low D coordinates of samples')
if peratom:
np.savetxt(foutput + "-atomic.coord", proj_atomic_all, fmt='%4.8f', header='low D coordinates of samples')
if output == 'xyz':
asapxyz.set_descriptors(proj, 'pca_coord')
if peratom:
asapxyz.set_atomic_descriptors(proj_atomic_all, 'pca_coord')
# remove the raw descriptors
if not keepraw:
asapxyz.remove_descriptors(fmat)
asapxyz.remove_atomic_descriptors(fmat)
asapxyz.write(foutput)
# color scheme
plotcolor, plotcolor_peratom, colorlabel, colorscale = set_color_function(fcolor, asapxyz, colorscol, 0, (peratom or plotatomic), projectatomic)
if plotatomic:
outfile = 'PCA_4_' + prefix + '-c-' + fcolor + '-plotatomic.png'
else:
outfile = 'PCA_4_' + prefix + '-c-' + fcolor + '.png'
fig_spec_dict = {
'outfile': outfile,
'show': False,
'title': None,
'xlabel': 'Principal Axis 1',
'ylabel': 'Principal Axis 2',
'xaxis': True, 'yaxis': True,
'remove_tick': False,
'rasterized': True,
'fontsize': 16,
'components':{
"first_p": {"type": 'scatter', 'clabel': colorlabel},
"second_p": {"type": 'annotate', 'adtext': adtext}
}
}
asap_plot = Plotters(fig_spec_dict)
asap_plot.plot(proj[::-1, [pc1, pc2]], plotcolor[::-1], [], tags)
if peratom or plotatomic and not projectatomic:
asap_plot.plot(proj_atomic_all[::-1, [pc1, pc2]], plotcolor_peratom[::-1],[],[])
plt.show()
# make tags
if ele == 14 or ele == 12:
n_atoms = 32
n_frames = get_nframes(fxyz_recal_om8, 'xyz')
tag0 = sum(np.array(n_frames).astype(int)) * n_atoms * [0]
n_frames = get_nframes(fxyz_recal_pv, 'xyz')
tag1 = sum(np.array(n_frames).astype(int)) * n_atoms * [1]
tags = tag0 + tag1
# tag0=(n_atoms*9 + n_atoms*250)*[0] + n_atoms*100*[1]
###########
########### The following is equivalent to above and have
###########
asapxyz = ASAPXYZ(fxyz)
reduce_dict = {}
reduce_dict["preprocessing"] = {"type": 'SCALE', 'parameter': None}
reduce_dict['skpca'] = {
"type": 'SPARSE_KPCA',
'parameter': {
"n_components": 3,
"n_sparse": -1, # no sparsification
# "scale":True,
"kernel": {
"first_kernel": {
"type": 'linear'
}
}
}
def main(fxyz, prefix):
"""
Test if computing descriptors is working.
Parameters
----------
fxyz: string giving location of xyz file
prefix: string giving the filename prefix
"""
# read frames
asapxyz = ASAPXYZ(fxyz, 1, False) # not periodic
peratom = True
tag = 'test'
soap_js = {
'soap1': {
'type': 'SOAP',
'cutoff': 2.0,
'n': 2,
'l': 2,
'atom_gaussian_width': 0.2,
'rbf': 'gto',
'crossover': False
}
}
acsf_js = {
'acsf1': {
'type': 'ACSF',
'cutoff': 2.0,
'g2_params': [[1, 1], [1, 2], [1, 3]],
'g4_params': [[1, 1, 1], [1, 2, 1], [1, 1, -1], [1, 2, -1]]
}
}
k2_js = {
'lmbtr-k2': {
'type': 'LMBTR_K2',
'k2': {
"geometry": {
"function": "distance"
},
"grid": {
"min": 0,
"max": 2,
"n": 10,
"sigma": 0.1
},
"weighting": {
"function": "exponential",
"scale": 0.5,
"cutoff": 1e-3
}
},
'periodic': False,
'normalization': "l2_each"
}
}
kernel_js = {}
kernel_js['k1'] = {
'reducer_type': 'moment_average',
'zeta': 2,
'element_wise': False
}
kernel_js['k2'] = {'reducer_type': 'sum', 'element_wise': True}
desc_spec_js = {
'test_cm': {
'type': "CM"
},
'test_soap': {
'atomic_descriptor': soap_js,
'reducer_function': kernel_js
},
'test_acsf': {
'atomic_descriptor': acsf_js,
'reducer_function': kernel_js
},
'test_k2': {
'atomic_descriptor': k2_js,
'reducer_function': kernel_js
}
}
# compute the descripitors
asapxyz.compute_global_descriptors(desc_spec_js, [], peratom, tag)
asapxyz.write_computed_descriptors(prefix, ['test_cm', 'test_soap'], [0])
asapxyz.write(prefix)
asapxyz.save_state(tag)
def main(fmat, kmat, fxyz, ftags, prefix, fcolor, colorscol, dimension, pc1, pc2, algorithm, projectatomic, adtext):
"""
Parameters
----------
fmat: Location of descriptor matrix file or name of the tags in ase xyz file. You can use gen_descriptors.py to compute it.
kmat: Location of the kernel matrix.
fxyz: Location of xyz file for reading the properties.
ftags: Location of tags for the first M samples
prefix: Filename prefix. Default is ASAP.
fcolor: Properties for all samples (N floats) used to color the scatter plot,[filename/rho/cluster]
colorscol: The column number of the properties used for the coloring. Starts from 0.
dimension: The number of principle components to keep
pc1: int, default is 0, which principle axis to plot the projection on
pc2: int, default is 1, which principle axis to plot the projection on
algorithm: the algorithm for density-based clustering options are: ([dbscan], [fdb])
projectatomic: build the projection using the (big) atomic descriptor matrix
adtext: Whether to adjust the text (True/False)
Returns
-------
cluster labels, PCA plots
"""
if fmat == 'none' and kmat == 'none':
raise ValueError('Must provide either the low-dimensional coordinates fmat or the kernel matrix kmat')
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, desc_atomic = asapxyz.get_descriptors(fmat, projectatomic)
if projectatomic:
desc = desc_atomic.copy()
else:
asapxyz = None
print("Did not provide the xyz file. We can only output descriptor matrix.")
output = 'matrix'
# we can also load the descriptor matrix from a standalone file
if os.path.isfile(fmat[0]):
try:
desc = np.genfromtxt(fmat[0], dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
if kmat != 'none':
try:
kNN = np.genfromtxt(kmat, dtype=float)
print("loaded kernal matrix", kmat, "with shape", np.shape(kNN))
desc = kerneltodis(kNN)
except:
raise ValueError('Cannot load the coordinates')
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")
ndict = len(tags)
# now we do the clustering
if algorithm == 'dbscan':
# we compute the characteristic bandwidth of the data
# first select a subset of structures (20)
sbs = np.random.choice(np.asarray(range(len(desc))), 50, replace=False)
# the characteristic bandwidth of the data
sigma_kij = np.percentile(cdist(desc[sbs], desc, 'euclidean'), 100*10./len(desc))
trainer = sklearn_DB(sigma_kij, 5, 'euclidean') # adjust the parameters here!
do_clustering = DBCluster(trainer)
do_clustering.fit(desc)
elif algorithm == 'fdb' or algorithm == 'FDB':
trainer = LAIO_DB()
do_clustering = DBCluster(trainer)
do_clustering.fit(desc)
else:
raise ValueError('Please select from fdb or dbscan')
print(do_clustering.pack())
#with open("clustering_results_4_" + prefix + ".json", 'w') as fp:
# json.dump(do_clustering.pack(), fp, cls=NpEncoder)
labels_db = do_clustering.get_cluster_labels()
n_clusters = do_clustering.get_n_cluster()
if asapxyz is not None and projectatomic:
asapxyz.set_atomic_descriptors(labels_db, 'cluster_label')
elif asapxyz is not None:
asapxyz.set_descriptors(labels_db, 'cluster_label')
# save
np.savetxt(prefix + "-cluster-label.dat", np.transpose([np.arange(len(labels_db)), labels_db]),
header='index cluster_label', fmt='%d %d')
if fmat != 'none':
pca = PCA(dimension, True)
proj = pca.fit_transform(desc)
elif kmat != 'none':
proj = KernelPCA(dimension).fit_transform(kNN)
# color scheme
if fcolor == 'cluster_label':
plotcolor = labels_db
colorlabel = 'cluster_label'
else:
if projectatomic:
_, plotcolor, colorlabel, colorscale = set_color_function(fcolor, asapxyz, colorscol, 0, True)
else:
plotcolor, colorlabel, colorscale = set_color_function(fcolor, asapxyz, colorscol, len(proj), False)
print(labels_db[::-1])
outfile = 'Clustering_4_' + prefix + '.png'
# make plot
fig_spec_dict = {
'outfile': outfile,
'show': True,
'title': None,
'xlabel': 'Principal Axis 1',
'ylabel': 'Principal Axis 2',
'xaxis': True, 'yaxis': True,
'remove_tick': False,
'rasterized': True,
'fontsize': 16,
'components':{
"first_p": {"type": 'scatter', 'clabel': colorlabel},
#"second_p": {"type": 'annotate', 'adtext': adtext},
"third_p": {"type": 'cluster', 'w_label': True, 'circle_size': 20}
}
}
asap_plot = Plotters(fig_spec_dict)
asap_plot.plot(proj[::-1, [pc1, pc2]], plotcolor[::-1], labels_db[::-1], []) #tags)
def main(fmat, fxyz, ftags, fcolor, colorscol, prefix, output, peratom, keepraw, scale, tsne_d, dim1, dim2, perplexity,
projectatomic, plotatomic, adtext):
"""
Parameters
----------
fmat: Location of descriptor matrix file or name of the tags in ase xyz file. You can use gen_descriptors.py to compute it.
fxyz: Location of xyz file for reading the properties.
ftags: Location of tags for the first M samples. Plot the tags on the t-SNE map.
fcolor: Location of a file or name of the tags in ase xyz file. It should contain properties for all samples (N floats) used to color the scatterplot'
colorscol: The column number of the properties used for the coloring. Starts from 0.
prefix: Filename prefix, default is ASAP
output: The format for output files ([xyz], [matrix]). Default is xyz.
peratom: Whether to output per atom t-SNE coordinates (True/False)
keepraw: Whether to keep the high dimensional descriptor when output is an xyz file (True/False)
scale: Scale the coordinates (True/False). Scaling highly recommanded.
tsne_d: Dimension of the embedded space.
dim1: Plot the projection along which principle axes
dim2: Plot the projection along which principle axes
projectatomic: build the projection using the (big) atomic descriptor matrix
perplexity: Perplexity setting for t-SNE: Typical values between 5 and 50.
plotatomic: Plot the PCA coordinates of all atomic environments (True/False)
adtext: Whether to adjust the texts (True/False)
Returns
-------
"""
foutput = prefix + "-pca-d" + str(tsne_d)
use_atomic_desc = (peratom or plotatomic or projectatomic)
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, desc_atomic = asapxyz.get_descriptors(fmat, use_atomic_desc)
if projectatomic: desc = desc_atomic.copy()
else:
asapxyz = None
print("Did not provide the xyz file. We can only output descriptor matrix.")
output = 'matrix'
# we can also load the descriptor matrix from a standalone file
if os.path.isfile(fmat[0]):
try:
desc = np.genfromtxt(fmat[0], dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
# sanity check
if len(desc) == 0:
raise ValueError('Please supply descriptor in a xyz file or a standlone descriptor matrix')
print("shape of the descriptor matrix: ", np.shape(desc), "number of descriptors: ", np.shape(desc[0]))
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")[:]
ndict = len(tags)
else:
tags = []
# scale & center
if scale:
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
print('Shape of descriptor matrix is {}'.format(desc.shape))
print(scaler.fit(desc))
desc = scaler.transform(desc) # normalizing the features
# fit t-SNE
if desc.shape[1] >= 50:
# pre-process with PCA if dim > 50
# suggested here: https://scikit-learn.org/stable/modules/generated/sklearn.manifold.TSNE.html
pca = PCA(n_components=50)
desc = pca.fit_transform(desc)
print('Shape of processed descriptor matrix after applying PCA is {}'.format(desc.shape))
tsne = TSNE(n_components=tsne_d, perplexity=perplexity)
proj = tsne.fit_transform(desc)
if peratom or plotatomic and not projectatomic:
raise NotImplementedError
#proj_atomic_all = tsne.transform(desc_atomic)
# save
if output == 'matrix':
np.savetxt(foutput + ".coord", proj, fmt='%4.8f', header='low D coordinates of samples')
if output == 'xyz':
if os.path.isfile(foutput + ".xyz"):
os.rename(foutput + ".xyz", "bck." + foutput + ".xyz")
asapxyz.set_descriptors(proj, 'pca_coord')
if peratom:
asapxyz.set_atomic_descriptors(proj_atomic_all, 'pca_coord')
# remove the raw descriptors
if not keepraw:
asapxyz.remove_descriptors(fmat)
asapxyz.remove_atomic_descriptors(fmat)
asapxyz.write(foutput)
# color scheme
plotcolor, plotcolor_peratom, colorlabel, colorscale = set_color_function(fcolor, asapxyz, colorscol, 0, (peratom or plotatomic), projectatomic)
# make plot
plot_styles.set_nice_font()
fig, ax = plt.subplots()
if plotatomic and not projectatomic:
# notice that we reverse the list of coordinates, in order to make the structures in the dictionary more obvious
fig, ax = plot_styles.plot_density_map(proj_atomic_all[::-1, [dim1, dim2]], plotcolor_peratom[::-1], fig, ax,
xlabel='Principal Axis ' + str(dim1),
ylabel='Principal Axis ' + str(dim2),
clabel=None, label=None,
xaxis=True, yaxis=True,
centers=None,
psize=None,
out_file=None,
title=None,
show=False, cmap='gnuplot',
remove_tick=False,
use_perc=False,
rasterized=True,
fontsize=15,
vmax=colorscale[1],
vmin=colorscale[0])
fig, ax = plot_styles.plot_density_map(proj[::-1, [dim1, dim2]], plotcolor[::-1], fig, ax,
xlabel='Principal Axis ' + str(dim1), ylabel='Principal Axis ' + str(dim2),
clabel=colorlabel, label=None,
xaxis=True, yaxis=True,
centers=None,
psize=None,
out_file=None,
title='t-SNE for: ' + prefix,
show=False, cmap='gnuplot',
remove_tick=False,
use_perc=False,
rasterized=True,
fontsize=15,
vmax=colorscale[1],
vmin=colorscale[0])
fig.set_size_inches(160.5, 80.5)
if ftags != 'none':
texts = []
for i in range(ndict):
if tags[i] != 'None' and tags[i] != 'none' and tags[i] != '':
ax.scatter(proj[i, dim1], proj[i, dim2], marker='^', c='black')
texts.append(ax.text(proj[i, dim1], proj[i, dim2], tags[i],
ha='center', va='center', fontsize=10, color='red'))
if adtext:
from adjustText import adjust_text
adjust_text(texts, on_basemap=True, # only_move={'points':'', 'text':'x'},
expand_text=(1.01, 1.05), expand_points=(1.01, 1.05),
force_text=(0.03, 0.5), force_points=(0.01, 0.25),
ax=ax, precision=0.01,
arrowprops=dict(arrowstyle="-", color='black', lw=1, alpha=0.8))
plt.show()
if plotatomic:
fig.savefig('t-SNE_4_' + prefix + '-c-' + fcolor + '-plotatomic.png')
else:
fig.savefig('t-SNE_4_' + prefix + '-c-' + fcolor + '.png')
def main(fxyz, fy, prefix, nkeep, algorithm, fmat, fkde, reweight_lambda):
"""
Select frames from the supplied xyz file (fxyz) using one of the following algorithms:
1. random: random selection
2. fps: farthest point sampling selection. Need to supply a kernel matrix or descriptor matrix using -fmat
3. sortmin/sortmax: select the frames with the largest/smallest value. Need to supply the vector of properties using
-fy
4. CUR decomposition
5. Reweight according to the re-weighted distribution exp(-f/\lambda),
where exp(-f) is the precomputed kernel density estimation of the original samples.
Parameters
----------
fxyz: Path to xyz file.
fy: Path to the list of properties (N floats) or name of the tags in ase xyz file
prefix: Filename prefix, default is ASAP
nkeep: The number of representative samples to select
algorithm: 'the algorithm for selecting frames ([random], [fps], [sort], [reweight])')
fmat: Location of descriptor or kernel matrix file. Needed if you select [fps].
You can use gen_kmat.py to compute it.
reweight_lambda: select samples according to the re-weighted distribution exp(-f/\lambda),
where exp(-f) is the kernel density estimation of the original samples.
"""
# read the xyz file
asapxyz = ASAPXYZ(fxyz)
nframes = asapxyz.get_num_frames()
if nkeep == 0:
nkeep = nframes
if fy != 'none':
y_all = []
try:
y_all = np.genfromtxt(fy, dtype=float)
except:
y_all = asapxyz.get_property(fy)
if len(y_all) != nframes:
raise ValueError(
'Length of the vector of properties is not the same as number of samples'
)
if algorithm == 'random' or algorithm == 'RANDOM':
idx = np.asarray(range(nframes))
sbs = np.random.choice(idx, nkeep, replace=False)
elif algorithm == 'sortmax' or algorithm == 'sortmin':
if fy == 'none':
raise ValueError(
'must supply the vector of properties for sorting')
idx = np.asarray(range(nframes))
if algorithm == 'sortmax':
sbs = [x for _, x in sorted(zip(y_all, idx))][:nkeep]
elif algorithm == 'sortmin':
sbs = [x for _, x in sorted(zip(y_all, idx))][nkeep:]
elif algorithm == 'fps' or algorithm == 'FPS' or algorithm == 'cur' or algorithm == 'CUR':
# for both algo we read in the descriptor matrix
desc, _ = asapxyz.get_descriptors(fmat)
if os.path.isfile(fmat):
try:
desc = np.genfromtxt(fmat, dtype=float)
except:
raise ValueError('Cannot load the kernel matrix')
print("shape of the descriptor matrix: ", np.shape(desc),
"number of descriptors: ", np.shape(desc[0]))
# FPS
if algorithm == 'fps' or algorithm == 'FPS':
sbs, dmax_remain = fps(desc, nkeep, 0)
print("Making farthest point sampling selection")
np.savetxt(prefix + "-" + algorithm + "-n-" + str(nkeep) +
'.error',
dmax_remain,
fmt='%4.8f',
header='the maximum remaining distance in FPS')
# CUR decomposition
if algorithm == 'cur' or algorithm == 'CUR':
desc = np.asmatrix(desc)
cov = np.dot(desc, desc.T)
print("Making CUR selection")
print("shape of the covariance matrix:", np.shape(cov))
sbs, rcov_error = CUR_deterministic(cov, nkeep)
np.savetxt(prefix + "-" + algorithm + "-n-" + str(nkeep) +
'.error',
rcov_error,
fmt='%4.8f',
header='the remaining error of the covariance matrix')
elif algorithm == 'reweight':
if os.path.isfile(fkde):
try:
logkde = np.genfromtxt(fkde, dtype=float)[:, 1]
except:
raise IOError(
'Cannot load the (log of) kernel density for each sample')
if len(logkde) != nframes:
raise ValueError(
'mismatch of number of frames and kernel densities')
else:
raise ValueError(
'must suply the (log of) kernel density for each sample')
new_kde = np.zeros(nframes)
for i in range(nframes):
new_kde[i] = np.exp(logkde[i] / reweight_lambda) / np.exp(
logkde[i])
# compute the normalization factor so we expect to select n samples in the end
normalization = nkeep / np.sum(new_kde)
new_kde *= normalization
sbs = []
randomchoice = np.random.rand(nframes)
for i in range(nframes):
if randomchoice[i] < new_kde[i]:
sbs.append(i)
algorithm = algorithm + "-lambda-" + str(reweight_lambda)
# save
selection = np.zeros(nframes, dtype=int)
for i in sbs:
selection[i] = 1
np.savetxt(prefix + "-" + algorithm + "-n-" + str(nkeep) + '.index',
selection,
fmt='%d')
if fy != 'none':
np.savetxt(prefix + "-" + algorithm + "-n-" + str(nkeep) + '-' + fy,
np.asarray(y_all)[sbs],
fmt='%4.8f')
asapxyz.write(prefix + "-" + algorithm + "-n-" + str(nkeep), sbs)
def main():
"""
Test if dimensionality reduction is working.
Parameters
----------
fxyz: string giving location of xyz file
prefix: string giving the filename prefix
"""
fxyz = os.path.join(os.path.split(__file__)[0], 'small_molecules-SOAP.xyz')
fmat = ['SOAP-n4-l3-c1.9-g0.23']
fcolor = 'dft_formation_energy_per_atom_in_eV'
pca_d = 10
prefix = "test-dimensionality-reduction"
foutput = prefix + "-pca-d" + str(pca_d)
# try to read the xyz file
asapxyz = ASAPXYZ(fxyz)
desc, _ = asapxyz.get_descriptors(fmat, False)
print(desc)
"""
reduce_dict = { "pca":
{"type": 'PCA', 'parameter':{"n_components": pca_d, "scalecenter": scale}}
}
reduce_dict = {
"preprocessing": {"type": 'SCALE', 'parameter': None},
"umap":
{"type": 'UMAP', 'parameter':{"n_components": pca_d, "n_neighbors": 10}}
}
reduce_dict = {
"reduce1_pca": {"type": 'PCA', 'parameter':{"n_components": 20, "scalecenter":True}},
"reduce2_tsne": {"type": 'TSNE', 'parameter': {"n_components": 2, "perplexity":20}}
}
"""
reduce_dict = {
"preprocessing": {"type": 'SCALE', 'parameter': None},
"skpca":
{"type": 'SPARSE_KPCA',
'parameter':{"n_components": pca_d,
"kernel": {"first_kernel": {"type": 'linear', "normalize": True}}
}
}
}
dreducer = Dimension_Reducers(reduce_dict)
proj = dreducer.fit_transform(desc)
# save
asapxyz.set_descriptors(proj, 'pca_coord')
asapxyz.write(foutput)
# color scheme
plotcolor, plotcolor_peratom, colorlabel, colorscale = set_color_function(fcolor, asapxyz)
outfile = 'PCA_4_' + prefix + '-c-' + fcolor + '.png'
fig_spec_dict = {
'outfile': outfile,
'show': False,
'title': None,
'xlabel': 'Principal Axis 1',
'ylabel': 'Principal Axis 2',
'xaxis': True, 'yaxis': True,
'remove_tick': False,
'rasterized': True,
'fontsize': 16,
'components':{
"first_p": {"type": 'scatter', 'clabel': colorlabel},
"second_p": {"type": 'annotate', 'adtext': False}
}
}
asap_plot = Plotters(fig_spec_dict)
asap_plot.plot(proj[::-1, [0, 1]], plotcolor[::-1], [], [])
plt.show()
def main(fmat, fxyz, ftags, fcolor, colorscol, prefix, output, peratom,
keepraw, sparse_mode, n_sparse, power, kpca_d, pc1, pc2,
projectatomic, plotatomic, adjusttext):
"""
Parameters
----------
fmat: Location of descriptor matrix file or name of the tags in ase xyz file. You can use gen_descriptors.py to compute it.
fxyz: Location of xyz file for reading the properties.
ftags: Location of tags for the first M samples. Plot the tags on the (k)PCA map.
fcolor: Location of a file or name of the tags in ase xyz file. It should contain properties for all samples (N floats) used to color the scatterplot'
colorscol: The column number of the properties used for the coloring. Starts from 0.
prefix: Filename prefix, default is ASAP
output: The format for output files ([xyz], [matrix]). Default is xyz.
peratom: Whether to output per atom pca coordinates (True/False)
keepraw: Whether to keep the high dimensional descriptor when output is an xyz file (True/False)
n_sparse: number of representative samples, default is 5% of the data
power: use polynomial kernel function of degree n.
kpca_d: Number of the principle components to keep
pc1: Plot the projection along which principle axes
pc2: Plot the projection along which principle axes
projectatomic: build the projection using the (big) atomic descriptor matrix
plotatomic: Plot the PCA coordinates of all atomic environments (True/False)
adtext: Whether to adjust the texts (True/False)
Returns
-------
"""
use_atomic_desc = (peratom or plotatomic or projectatomic)
# try to read the xyz file
if fxyz != 'none':
asapxyz = ASAPXYZ(fxyz)
desc, desc_atomic = asapxyz.get_descriptors(fmat, use_atomic_desc)
if projectatomic: desc = desc_atomic.copy()
else:
asapxyz = None
print(
"Did not provide the xyz file. We can only output descriptor matrix."
)
output = 'matrix'
# we can also load the descriptor matrix from a standalone file
if os.path.isfile(fmat[0]):
try:
desc = np.genfromtxt(fmat[0], dtype=float)
print("loaded the descriptor matrix from file: ", fmat)
except:
raise ValueError('Cannot load the descriptor matrix from file')
# sanity check
if len(desc) == 0:
raise ValueError(
'Please supply descriptor in a xyz file or a standlone descriptor matrix'
)
print("shape of the descriptor matrix: ", np.shape(desc),
"number of descriptors: ", np.shape(desc[0]))
if ftags != 'none':
tags = np.loadtxt(ftags, dtype="str")[:]
ndict = len(tags)
else:
tags = []
# sparsification
n_sample = len(desc)
# set default value of n_sparse
if n_sparse == 0:
n_sparse = max(10, n_sample // 20)
# sparsification
if n_sparse >= n_sample:
print(
"the number of representative structure is too large, please select n < ",
n_sample)
elif n_sample > 0:
if sparse_mode == 'fps' or sparse_mode == 'FPS':
ifps, _ = fps(desc, n_sparse, 0)
elif sparse_mode == 'cur' or sparse_mode == 'CUR':
cov = np.dot(np.asmatrix(desc), np.asmatrix(desc).T)
ifps, _ = CUR_deterministic(cov, n_sparse)
else:
raise ValueError('Cannot find the specified sparsification mode')
else:
print("Not using any sparsification")
ifps = np.range(n_sample)
k_spec = {
'k0': {
"type": "cosine"
}
} #{ 'k1': {"type": "polynomial", "d": power}}
k_transform = Descriptors_to_Kernels(k_spec)
kNN = k_transform.compute(desc[ifps])
kMN = k_transform.compute(desc, desc[ifps])
print("Shape of the kNN matrix: ", np.shape(kNN),
", and shape of the kMN matrix:", np.shape(kMN))
# main thing
kpca = KernelPCA(kpca_d)
kpca.fit(kNN)
proj = kpca.transform(kMN)
if peratom or plotatomic and not projectatomic:
kNT = np.power(np.dot(desc_atomic[:], desc[ifps].T), power)
proj_atomic_all = kpca.transform(kNT)
# save
if output == 'matrix':
np.savetxt(prefix + "-kpca-d" + str(kpca_d) + ".coord",
proj,
fmt='%4.8f',
header='low D coordinates of samples')
elif output == 'xyz':
if os.path.isfile(foutput + ".xyz"):
os.rename(foutput + ".xyz", "bck." + foutput + ".xyz")
asapxyz.set_descriptors(proj, 'kpca_coord')
asapxyz.write(foutput)
# color scheme
plotcolor, plotcolor_peratom, colorlabel, colorscale = set_color_function(
fcolor, asapxyz, colorscol, 0, (peratom or plotatomic), projectatomic)
# make plot
if plotatomic:
outfile = 'KPCA_4_' + prefix + '-c-' + fcolor + '-plotatomic.png'
else:
outfile = 'KPCA_4_' + prefix + '-c-' + fcolor + '.png'
fig_spec_dict = {
'outfile': outfile,
'show': False,
'title': None,
'xlabel': 'Principal Axis 1',
'ylabel': 'Principal Axis 2',
'xaxis': True,
'yaxis': True,
'remove_tick': False,
'rasterized': True,
'fontsize': 16,
'components': {
"first_p": {
"type": 'scatter',
'clabel': colorlabel
},
"second_p": {
"type": 'annotate',
'adtext': adjusttext
}
}
}
asap_plot = Plotters(fig_spec_dict)
asap_plot.plot(proj[::-1, [pc1, pc2]], plotcolor[::-1], [], tags)
if peratom or plotatomic and not projectatomic:
asap_plot.plot(proj_atomic_all[::-1, [pc1, pc2]],
plotcolor_peratom[::-1], [], [])
plt.show()