def mk2set(ind ,randseed):
p,n,psamp,nsamp,ptrain,ntrain = main.alt_lc_get_graphs(randseed)
ba.dumpfile(p,f"{foldername}/ptest_{ind}.pick")
ba.dumpfile(n,f"{foldername}/ntest_{ind}.pick")
rdk.nx_to_moses(ptrain,f"{foldername}/ptrain_{ind}.csv")
rdk.nx_to_moses(ntrain,f"{foldername}/ntrain_{ind}.csv")
def getnx(fname):
cachename = fname + ".cache"
if os.path.isfile(cachename):
return ba.loadfile(cachename)
with gzip.open(fname, 'rb') as fi:
smiles = fi.read()
atomz = list(
rut.smiles_strings_to_nx(
[line.split()[1] for line in smiles.split(b'\n')[:-1]]))
random.seed(123)
random.shuffle(atomz)
ba.dumpfile(atomz, cachename)
return atomz
def loadsmi(fname, randseed=123):
'''can we load from cache?'''
cachename = fname + ".cache"
if os.path.isfile(cachename):
graphs = ba.loadfile(cachename)
else:
g = list(rut.smi_to_nx(fname))
graphs = lu.pre_process(g)
ba.dumpfile(graphs, cachename)
random.seed(randseed)
random.shuffle(graphs)
return graphs
def getnx(fname, randseed=123):
'''can we load from cache?'''
cachename = fname + ".cache"
if os.path.isfile(cachename):
graphs = ba.loadfile(cachename)
else:
'''if not load normaly and write a cache'''
with gzip.open(fname, 'rb') as fi:
smiles = fi.read()
graphs = list(
rut.smiles_strings_to_nx(
[line.split()[1] for line in smiles.split(b'\n')[:-1]]))
graphs = lu.pre_process(graphs)
ba.dumpfile(graphs, cachename)
'''shuffle and return'''
random.seed(randseed)
random.shuffle(graphs)
return graphs
def format_abc(a, b, c, sav='res.pickle'):
cm = [np.mean(x) for x in list(zip(*a))]
om = [np.mean(x) for x in list(zip(*b))]
gm = [np.mean(x) for x in list(zip(*c))]
cs = [np.std(x) for x in list(zip(*a))]
os = [np.std(x) for x in list(zip(*b))]
gs = [np.std(x) for x in list(zip(*c))]
logger.log(51, f'combined {cm}')
logger.log(51, f'originals {om}')
logger.log(51, f'generated {gm}')
logger.log(41, f'combined{cs}')
logger.log(41, f'originals only{os}')
logger.log(41, f'generated only{gs}')
ts = np.array(args.trainsizes)
gen = np.array([
e * ((args.n_steps - args.burnin) // args.emit + 1)
for e in args.trainsizes
])
ba.dumpfile([(ts + gen, ts, gen), (cm, om, gm), (cs, os, gs), (a, b, c)],
sav)
def get_params_punk():
trve = lambda pp: [i for d in [pp.a, pp.b] for i in d.obs['true'].values]
dnames = "human1 human2 human3 human4 smartseq2 celseq2 celseq".split()
loader = lambda x, seed: load.loadgruen_single(
f"../data/punk/{x}", subsample=sampnum, seed=seed)
return dnames, loader, trve
if __name__ == "__main__":
task, t2, rep = map(int, sys.argv[1].strip().split(' '))
dnames, loader, trve = get_params_100()
other = dnames[t2]
self = dnames[task]
result = get_score(self, other, loader, trve, seed=rep)
print("res: ", result)
ba.dumpfile(result, "res/" + sys.argv[1].replace(" ", '_'))
print("all good")
# use median instead of mean! TODO
# add error baro to plot -> quantiles! plot points! fit forrcoeff linear
# check the seed value for subsampling
# look at 100
def res(indices, reps):
print(dnames)
for i in range(indices):
indexrepeats = np.array(
[ba.loadfile(f"res/{i}_{r}") for r in range(reps)])
print(indexrepeats.mean(axis=0).tolist())
def save(self):
ba.dumpfile(self.data, self.fname)
ba.jdumpfile(self.data_int, self.fname_int)