def prepare_neutron_data(infpn, output_preview=False):
# load the neutron dataframe
ndata = pd.read_csv(infpn)
# select only neutrons produced by the interaction
# proton + Be9 --> proton + neutron + Be8
ndata = ndata[(ndata.particle_name == 'neutron') & (
ndata.equation == 'proton + Be9 --> proton + neutron + Be8')]
if output_preview:
ndata.to_csv(
os.path.join(data_utilities.get_git_root(__file__),
'data/p_on_be/pandas/be8_neutrons.csv'))
# track_id is categorical. Might be useful, but skip for now.
# Be9 is also categorical and skipped for now.
sel_cols = ['posx', 'posy', 'posz', 'px', 'py', 'pz', 'ke']
return ndata[sel_cols]
print(vargroupname, 'finished!')
def simplify_process_categories(self, eqn_str):
'''
Since it is quite cumbersome to track processes in which neutrons are produced by secondary particles,
they are simply grouped into one category called "neutrons produced by secondaries".
'''
_, daughters = data_utilities.parse_equation(eqn_str)
if not 'neutron' in daughters:
return 'neutrons produced by secondaries'
return eqn_str
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('-i', '--input_file', type=str, default=os.path.join(data_utilities.get_git_root(__file__), 'data/p_on_be/pandas/neutron_preview.csv'))
args = parser.parse_args()
infpn = args.input_file
# configure pandas to print all rows
# pd.set_option('display.max_rows', None)
# load dataframe
ndata = neutron_from_target(infpn)
# print(ndata.df_neutron)
# create output folder
outdir = 'plots'
if not os.path.exists(outdir):
os.makedirs(outdir)
# track_id is categorical. Might be useful, but skip for now.
# Be9 is also categorical and skipped for now.
sel_cols = ['posx', 'posy', 'posz', 'px', 'py', 'pz', 'ke']
return ndata[sel_cols]
if __name__ == '__main__':
# command line arguments
parser = argparse.ArgumentParser()
parser.add_argument('-i',
'--input_file',
type=str,
default=os.path.join(
data_utilities.get_git_root(__file__),
'data/p_on_be/pandas/neutron_preview.csv'))
args = parser.parse_args()
infpn = args.input_file
# preprocess data
X = prepare_neutron_data(infpn, output_preview=True)
print('Before standardization')
print(X)
standarized_X = StandardScaler().fit_transform(X)
print('After standardization')
print(standarized_X)
# Create figure
fig = plt.figure(figsize=(12, 8))
fig.suptitle('Manifold Learning comparison', fontsize=14)
cosine_edges = list(cosines)
for cos in list(cosines):
edge = cos - min_cos_interval
if not edge in cosine_edges:
cosine_edges.append(cos - min_cos_interval)
return sorted(np.round(np.arccos(np.array(cosine_edges)) * 180 / np.pi, 4))
if __name__ == '__main__':
# command line arguments
parser = argparse.ArgumentParser()
parser.add_argument('-i',
'--input_file',
type=str,
default=os.path.join(
data_utilities.get_git_root(__file__),
'data/p_on_be/pandas/neutron_preview.csv'))
parser.add_argument('--equal_angle_bin', action='store_true')
args = parser.parse_args()
infpn = args.input_file
# load data from file into a dataframe
df_neutron = pd.read_csv(infpn)
df_neutron = df_neutron[df_neutron.particle_name == 'neutron']
df_neutron['theta_p_degree'] = df_neutron.theta_p * 180 / np.pi
if args.equal_angle_bin:
angle_edges = [
0., 20., 40., 60., 80., 90., 100., 120., 140., 160., 180.
]
else: