def from_csv():
data = open('data.csv')
start = True
for line in data:
if start:
start = False
continue
split = line.split(',')
smiles = split[0]
print(list(tokenize_smiles(smiles)))
print('end')
def get_tokens_dict():
st_dict = {}
with open("canonized_smiles.txt", 'r') as f:
smiles = [line.rstrip('\n') for line in f]
index = 2
for smile in smiles:
tokenized_smile = list(tokenize_smiles(smile))
for token in tokenized_smile:
if not token in st_dict:
st_dict[token] = index
index += 1
return st_dict
def get_smiles_as_vectors():
with open("canonized_smiles.txt", 'r') as f:
smiles = [line.rstrip('\n') for line in f]
tokens_dict = get_tokens_dict()
r = []
for smile in smiles:
vector = []
tokens = list(tokenize_smiles(smile))
for token in tokens:
vector.append(tokens_dict[token])
r.append(vector)
return r
def get_smiles_numerical_vectors_dict():
with open("canonized_smiles.txt", 'r') as f:
smiles = [line.rstrip('\n') for line in f]
tokens_dict = get_tokens_dict()
d = {}
for smile in smiles:
vector = []
tokens = list(tokenize_smiles(smile))
for token in tokens:
vector.append(tokens_dict[token])
d[smile] = vector
return d
def from_npy():
smiles = np.loadtxt("canonized_data.npy", dtype=str)
for smile in smiles:
print(list(tokenize_smiles(smile)))