"""
Function to read a SLUG2 integrated_sn file.
"""
import numpy as np
from collections import namedtuple
from collections import defaultdict
import struct
import re
from .slug_open import slug_open
[docs]def read_integrated_sn(model_name, output_dir=None, fmt=None,
verbose=False, read_info=None):
"""
Function to read a SLUG2 integrated_sn file.
Parameters
model_name : string
The name of the model to be read
output_dir : string
The directory where the SLUG2 output is located; if set to None,
the current directory is searched, followed by the SLUG_DIR
directory if that environment variable is set
fmt : 'txt' | 'ascii' | 'bin' | 'binary' | 'fits' | 'fits2'
Format for the file to be read. If one of these is set, the
function will only attempt to open ASCII-('txt' or 'ascii'),
binary ('bin' or 'binary'), or FITS ('fits' or 'fits2')
formatted output, ending in .txt., .bin, or .fits,
respectively. If set to None, the code will try to open
ASCII files first, then if it fails try binary files, and if
it fails again try FITS files.
verbose : bool
If True, verbose output is printed as code runs
read_info : dict
On return, this dict will contain the keys 'fname' and
'format', giving the name of the file read and the format it
was in; 'format' will be one of 'ascii', 'binary', or 'fits'
Returns
A namedtuple containing the following fields:
time : array, shape (N_times) or shape (N_trials)
Times at which data are output; shape is either N_times (if
the run was done with fixed output times) or N_trials (if
the run was done with random output times)
tot_sn : array, shape (N_cluster)
total number of supernovae produced by each cluster up to
the indicated time
stoch_sn : array, shape (N_cluster)
total number of supernovae produced by stars being treated
stochastically in each cluster up to the indicated time
Raises
IOError, if no sn file can be opened
"""
# Open file
fp, fname = slug_open(model_name+"_integrated_sn",
output_dir=output_dir,
fmt=fmt)
# See if this file is a checkpoint file
if len(re.findall('_chk\d\d\d\d', model_name)) != 0:
checkpoint = True
else:
checkpoint = False
# Print status and record
if verbose:
print("Reading integrated supernovae for model "+model_name)
if read_info is not None:
read_info['fname'] = fname
# Prepare lists to hold data
time = []
tot_sn = []
stoch_sn = []
# Read data
if fname.endswith('.txt'):
# ASCII mode
if read_info is not None:
read_info['format'] = 'ascii'
# If this is a checkpoint file, skip the line stating how many
# trials it contains; this line is not guaranteed to be
# accurate, and is intended for the C++ code, not for us
if checkpoint:
fp.readline()
# Read the first three header lines
hdr = fp.readline()
fp.readline()
fp.readline()
# Read data
trial = []
trialptr = 0
for entry in fp:
if entry[:3] == '---':
trialptr = trialptr+1
continue # Skip separator lines
trial.append(trialptr)
data = entry.split()
time.append(float(data[0]))
tot_sn.append(float(data[1]))
stoch_sn.append(int(data[2]))
elif fname.endswith('.bin'):
# Binary mode
if read_info is not None:
read_info['format'] = 'binary'
# If this is a checkpoint, skip the bytes specifying how many
# trials we have; this is inteded for the C++ code, not for
# us, since we will determine that on our own
if checkpoint:
data = fp.read(struct.calcsize('i'))
# Suck file into memory
data = fp.read()
# Interpret data
datstr = 'LddL'
nentry = len(data)//struct.calcsize(datstr)
data_list = struct.unpack(datstr*nentry, data)
# Stick data into correctly-named lists
trial = data_list[0::4]
time = data_list[1::4]
tot_sn = data_list[2::4]
stoch_sn = data_list[3::4]
elif fname.endswith('fits'):
# FITS mode
if read_info is not None:
read_info['format'] = 'fits'
trial = fp[1].data.field('Trial')
time = fp[1].data.field('Time')
tot_sn = fp[1].data.field('TotSN')
stoch_sn = fp[1].data.field('StochSN')
# Close file
fp.close()
# Convert lists to arrays
trial = np.array(trial)
time = np.array(time)
tot_sn = np.array(tot_sn)
stoch_sn = np.array(stoch_sn, dtype='ulonglong')
# Figure out if we have a number of trials with identical times,
# indicating fixed output times, or if each trial has random times;
# reshape time array appropriately
ntrial = len(np.unique(trial))
ntime = len(time)//ntrial
if ntime != len(time):
if np.amin(time[:ntime] == time[ntime:2*ntime]):
time = time[:ntime]
# Prune / reshape the output arrays
tot_sn = np.transpose(tot_sn.reshape(ntrial, ntime))
stoch_sn = np.transpose(stoch_sn.reshape(ntrial, ntime))
# Build the namedtuple to hold output
out_type = namedtuple('integrated_sn',
['time', 'tot_sn', 'stoch_sn'])
out = out_type(time, tot_sn, stoch_sn)
# Return
return out