'--nsamples',
type=int,
default=1024,
help='Number of posterior samples to average (only used for DTSRBayes)'
)
argparser.add_argument(
'-M',
'--mode',
type=str,
default=None,
help=
'Predict mode ("response" or "loglik") or default None, which does both'
)
args, unknown = argparser.parse_known_args()
p = Config(args.config_path)
if len(args.models) > 0:
models = args.models
else:
models = p.model_list[:]
run_baseline = False
run_dtsr = False
for m in models:
if not run_baseline and m.startswith('LM') or m.startswith('GAM'):
run_baseline = True
elif not run_dtsr and m.startswith('DTSR'):
run_dtsr = True
dtsr_formula_list = [
Formula(p.models[m]['formula']) for m in p.model_list
table_str += '\\end{tabular}\n\\end{table}'
print(table_str)
if __name__ == '__main__':
argparser = argparse.ArgumentParser('''
Generates a LaTeX table of beta summaries for one or more DTSR models
''')
argparser.add_argument('config', help='Path to config file defining models')
argparser.add_argument('-m', '--models', nargs='+', default=[], help='Folder name(s) containing model eval summaries (if blank applies to all models described in config file).')
argparser.add_argument('-n', '--names', nargs='*', default=[], help='Model names to print in table (must be omitted or same length as --models).')
args = argparser.parse_args()
p = Config(args.config)
if len(args.models) == 0:
models = p.model_list[:]
else:
models = args.models[:]
if len(args.names) == 0:
names = p.model_list[:]
else:
assert len(args.names) == len(models), 'Length mismatch between number of models and number of model names'
names = args.names
beta_summaries = {}
beta_names = set()
for i in range(len(models)):
argparser = argparse.ArgumentParser('''
Generates plots of fitted spillover estimates for "fullS" baseline models, visualizing those models' estimates of temporal diffusion.
''')
argparser.add_argument('paths', nargs='+', help='Paths to config files defining baseline models to plot')
argparser.add_argument('-m', '--models', nargs='*', default=None, help='List of models to plot. If none specified, plots all relevant models in the config files')
argparser.add_argument('-x', '--x', type=float, default=6, help='Width of plots in inches')
argparser.add_argument('-X', '--xlab', type=str, default=None, help='x-axis label (if default -- None -- no label)')
argparser.add_argument('-y', '--y', type=float, default=4, help='Height of plots in inches')
argparser.add_argument('-Y', '--ylab', type=str, default=None, help='y-axis label (if default -- None -- no label)')
argparser.add_argument('-c', '--cmap', type=str, default='gist_rainbow', help='Name of matplotlib colormap library to use for curves')
argparser.add_argument('-l', '--nolegend', action='store_true', help='Omit legend from figure')
args = argparser.parse_args()
configs = []
for c in args.paths:
configs.append(Config(c))
if args.models is None:
models = []
else:
models = [x for x in args.models if ('LMfullS' in x or 'LMEfullS' in x)]
assert len(models > 0), 'No valid models in script call. Models must be linear fullS models.'
cm = plt.get_cmap(args.cmap)
legend = not args.nolegend
for p in configs:
if len(models) == 0:
models = [x for x in p.model_list[:] if ('LMfullS' in x or 'LMEfullS' in x)]
for name in models:
dirpath = p.logdir + '/' + name + '/'
type=str,
default='gist_rainbow',
help='Name of matplotlib colormap library to use for curves')
argparser.add_argument('-l',
'--nolegend',
action='store_true',
help='Omit legend from figure')
argparser.add_argument(
'-t',
'--transparent_background',
action='store_true',
help='Use transparent background (otherwise white background)')
args, unknown = argparser.parse_known_args()
for path in args.paths:
p = Config(path)
if not p.use_gpu_if_available:
os.environ['CUDA_VISIBLE_DEVICES'] = '-1'
if len(args.models) > 0:
models = [
m for m in args.models
if (m in p.model_list and m.startswith('DTSR'))
]
else:
models = p.model_list[:]
prefix = args.prefix
if prefix != '':
prefix += '_'
n_time_units = args.ntimeunits
default=[],
help='List of paths to config files.')
argparser.add_argument(
'-p',
'--partition',
type=str,
default='dev',
help='Name of partition to use (one of "train", "dev", "test")')
args, unknown = argparser.parse_known_args()
losses = {}
select = {}
systems = set()
for task in args.tasks:
p = Config(task)
models = p.model_list[:]
systems = systems.union(
set([model2name(x.split('_')[0].strip()) for x in models]))
for m in models:
if os.path.exists(p.outdir + '/' + m + '/%s_losses_%s.txt' %
(p.loss, args.partition)):
sys_name, trial_name = m.split('_')
sys_name = model2name(sys_name)
if sys_name.startswith('LME') and trial_name == 'noRE':
l = pd.read_csv(p.outdir + '/' + m + '/%s_losses_%s.txt' %
(p.loss, args.partition),
sep=' ',
header=None,
skipinitialspace=True)
else: