def write_result_to(file_path, true_keywords, pred_keywords):
print("Writing true and predicted keywords to file {}...".format(file_path))
with open(file_path, "w", encoding="utf-8") as file:
for i in range(len(true_keywords)):
true_keyword = utils.list_to_str(true_keywords[i]).lower() # 把labels中的英文字母全部换成小写
pred_keyword = utils.list_to_str(pred_keywords[i])
pairs = true_keyword + "," + pred_keyword + "\n"
file.write(pairs)
print("Done.")
def print_stats(self):
logger.info("Training data stats: \n" +
"\t- intent examples: {} ({} distinct intents)\n".format(
len(self.intent_examples), len(self.intents)) +
"\t- Found intents: {}\n".format(
list_to_str(self.intents)) +
"\t- entity examples: {} ({} distinct entities)\n".format(
len(self.entity_examples), len(self.entities)) +
"\t- found entities: {}\n".format(
list_to_str(self.entities)))
def choose_designable_residues(pose,
focus_residues,
include_focus=True,
dshell=8.0,
pshell=12.0):
"""
Chooses a shell (for now, might make more sophisticated later) of residues
to design around the motif residue.
"""
focus_residue_selector =\
rosetta.core.select.residue_selector.ResidueIndexSelector(list_to_str(focus_residues))
designable_selector =\
rosetta.core.select.residue_selector.NeighborhoodResidueSelector(
focus_residue_selector, dshell,
include_focus_in_subset=include_focus
)
designable_not_selector =\
rosetta.core.select.residue_selector.NotResidueSelector(
designable_selector
)
packable_selector =\
rosetta.core.select.residue_selector.NeighborhoodResidueSelector(
focus_residue_selector, pshell,
include_focus_in_subset=include_focus
)
repack_only_selector =\
rosetta.core.select.residue_selector.AndResidueSelector(
designable_not_selector, packable_selector
)
design_residues = designable_selector.apply(pose)
repack_residues = repack_only_selector.apply(pose)
return design_residues, repack_residues
def cmd_afk(self, user, away_msg):
"""Marks a user afk
user: username of the user who issued the command, string
away_msg: away msg to be set, list of strings
"""
away_msg = utils.list_to_str(away_msg)
afk_users = []
with open("afk_users.csv", 'r') as f:
afk_users.extend(csv.reader(f))
self.debug_print(afk_users)
# Hacky fix for bug when no one is afk
if afk_users == []:
afk_users.append(['',''])
self.debug_print("Added empty row to afk_users")
set_afk = False
for row in afk_users:
if user == row[0]:
row[1] = away_msg
self.send_msg("You were already away, Your new afk message is: " + away_msg)
set_afk = True
if set_afk == False:
afk_users.append([user, away_msg])
self.send_msg("You are now afk.")
self.debug_print(afk_users)
with open("afk_users.csv", 'w') as f:
csv.writer(f).writerows(afk_users)
def compile(self):
from linkify_it_re import linkifyItRe
# todo 这个class 方法如何访问到类的私有属性
re_dict = self.re = linkifyItRe(self._opts)
tlds = self._tlds
self.onCompile()
if not self._tlds_replaced:
tlds_2ch_src_re = 'a[cdefgilmnoqrstuwxz]|b[abdefghijmnorstvwyz]|c[acdfghiklmnoruvwxyz]|d[ejkmoz]|e[cegrstu]|f[ijkmor]|g[abdefghilmnpqrstuwy]|h[kmnrtu]|i[delmnoqrst]|j[emop]|k[eghimnprwyz]|l[abcikrstuvy]|m[acdeghklmnopqrstuvwxyz]|n[acefgilopruz]|om|p[aefghklmnrstwy]|qa|r[eosuw]|s[abcdeghijklmnortuvxyz]|t[cdfghjklmnortvwz]|u[agksyz]|v[aceginu]|w[fs]|y[et]|z[amw]'
self._tlds.append(tlds_2ch_src_re)
tlds.append(re_dict['src_xn'])
re_dict['src_tlds'] = list_to_str(tlds, '|')
def untpl(tpl):
return tpl.replace('%TLDS%', re_dict['src_tlds'])
# todo 编译为i
re_dict['emial_fuzzy'] = untpl(re_dict['tpl_email_fuzzy'])
re_dict['link_fuzzy'] = untpl(re_dict['tpl_link_fuzzy'])
re_dict['link_no_ip_fuzzy'] = untpl(re_dict['tpl_link_no_ip_fuzzy'])
re_dict['host_fuzzy_test'] = untpl(re_dict['tpl_host_fuzzy_test'])
# Compile each schema
aliased = []
self._compiled = {} # // Reset compiled data
def cmd_remind(self, args, user):
"""Sets a reminder
args[0]: user to be reminded
args[1]: relative time
args[2]: time unit (s/m/h/dD/wW/M/yY/)
"""
if len(args) < 3:
self.send_msg("You did not provide me with enough information :/ " +\
"Syntax: !remind [user] [number] [time unit (s/m/h/dD/wW/M/yY/)] [msg]")
return
user_to_remind = args[0]
rel_time = args[1]
time_unit = args[2]
msg = args[3:]
if user_to_remind == "me":
user_to_remind = user
# Get time in seconds since epoch
cur_time = time.time()
reminder_time = None
if utils.represents_int(rel_time) is False:
self.send_msg("I don't know when to remind you :( "+\
"Syntax: !remind [user] [number] [time unit (s/m/h/d/w/M/y/)] [msg]")
return
if time_unit not in "smhdDwWMyYD":
self.send_msg("Oof, I don't know how to interpret that time unit :S Use one of these: /m/h/d/w/M/y/")
return
if time_unit == 's':
reminder_time = cur_time + int(rel_time)
elif time_unit == 'm':
reminder_time = cur_time + (int(rel_time) * 60)
elif time_unit == 'h':
reminder_time = cur_time + (int(rel_time) * 3600)
elif time_unit == 'd' or time_unit == 'D':
reminder_time = cur_time + (int(rel_time) * 86400)
elif time_unit == 'w' or time_unit == 'W':
reminder_time = cur_time + (int(rel_time) * 604800)
elif time_unit == 'M':
reminder_time = cur_time + (int(rel_time) * 2592000)
elif time_unit == 'y' or time_unit == 'Y':
reminder_time = cur_time + (int(rel_time) * 31536000)
self.send_msg("Okay, I will remind " + user_to_remind + " on " + time.strftime("%A %d %B %H:%M:%S %Y", time.localtime(reminder_time)))
del args[1:2]
t = Timer(reminder_time - cur_time, self.remind_user, [user_to_remind, utils.list_to_str(msg)])
t.start();
def parse_recv_data(self, data):
if data.startswith("PING"):
self.sock_send_str(data.replace("PING", "PONG"))
if data[0] != ':':
pass
if (self.nickname + " :End of /MOTD") in data:
self.join_channel()
self.get_online_users()
if data.split(' ')[1] == "353":
self.debug_print("Parsing userlist!")
self.parse_userlist(data)
elif data.split(' ')[1] == "PRIVMSG":
user = data.split(':')[1].split('!')[0] # potiental crash point?
msg = data.split(' ')[3]
if msg.startswith(":!"):
command_and_args = self.parse_arguments(data)
command = command_and_args[0]
args = command_and_args[1]
self.debug_print(command_and_args)
if command in self.preset_text_cmds:
self.send_msg(self.preset_text_cmds[command])
if command == "!say":
self.cmd_say(args)
elif command == "!choose":
self.cmd_choose(args)
elif command == "!ascii":
self.cmd_ascii(args)
elif command == "!afk":
self.cmd_afk(self.get_sender(data), args)
elif command == "!back" or command == "!rug" or command == "!brak":
self.cmd_back(self.get_sender(data))
elif command == "!where":
self.cmd_where(args)
elif command == "!remind":
self.cmd_remind(args, user)
# Debug-only commands
if self.DEBUG:
if command == "!stop":
if self.get_sender(data) == "MrTijn":
print("Stopping...")
sys.exit(0)
elif command == "!send_raw":
print("!send_raw")
print(self.send_raw(utils.list_to_str(args)))
def cmd_ascii(self, msg):
"""Print msg in big ascii art letters
msg - list of strings
"""
# Convert msg to string
msg = utils.list_to_str(msg)
line1 = ""
line2 = ""
line3 = ""
for char in msg:
char = char.lower()
line1 += AsciiArt.characters[char][0]
line2 += AsciiArt.characters[char][1]
line3 += AsciiArt.characters[char][2]
self.send_msg(line1)
self.send_msg(line2)
self.send_msg(line3)
def setup_task_factory(pose,
designable_residue_selector,
repackable_residue_selector,
motif_dict=None,
extra_rotamers=True,
limit_aro_chi2=True,
layered_design=True,
designable_aa_types=None,
prepare_focus=False):
"""
Adapted from XingJie Pan's code at
[email protected]:Kortemme-Lab/local_protein_sequence_design.git:
local_protein_sequence_design/basic.py
motif_dict should have the resnum as the key for the single-letter restype,
ex. {38:'E'}
If you want to set up a task factory to only design the 'focus' residue,
pass it a motif dict and prepare_focus=True
"""
def list_to_str(l):
return ','.join(list(str(i) for i in l))
task_factory = rosetta.core.pack.task.TaskFactory()
if len(designable_residue_selector) > 0 and not prepare_focus:
racaa = rosetta.core.pack.task.operation.RestrictAbsentCanonicalAASRLT(
)
if designable_aa_types is None or\
len(list(compress(xrange(len(designable_residue_selector)),\
designable_residue_selector))) != len(designable_aa_types):
racaa.aas_to_keep('GAPVILMFYWSTKRDENQ') # No Cys or His
else:
racaa.aas_to_keep(designable_aa_types[i])
designable_operation = rosetta.core.pack.task.operation.OperateOnResidueSubset(
racaa, designable_residue_selector)
task_factory.push_back(designable_operation)
if motif_dict:
for resnum in motif_dict:
selector = rosetta.core.select.residue_selector.ResidueIndexSelector(
str(resnum))
print(selector)
racaa = rosetta.core.pack.task.operation.RestrictAbsentCanonicalAASRLT(
)
racaa.aas_to_keep(motif_dict[resnum])
motif_operation = rosetta.core.pack.task.operation.OperateOnResidueSubset(
racaa, selector)
task_factory.push_back(motif_operation)
if len(repackable_residue_selector) > 0: # This is always going to be true
# given a res selector; should alter this logic at some point
repackable_operation = rosetta.core.pack.task.operation.OperateOnResidueSubset(
rosetta.core.pack.task.operation.RestrictToRepackingRLT(),
repackable_residue_selector)
task_factory.push_back(repackable_operation)
if prepare_focus and len(designable_residue_selector) > 0:
repackable_operation = rosetta.core.pack.task.operation.OperateOnResidueSubset(
rosetta.core.pack.task.operation.RestrictToRepackingRLT(),
designable_residue_selector)
task_factory.push_back(repackable_operation)
natro_residues = [
i for i in range(1,
pose.size() + 1)
if (not (designable_residue_selector[i]
or repackable_residue_selector[i]) and i not in motif_dict)
]
if len(natro_residues) > 0:
natro_selector =\
rosetta.core.select.residue_selector.ResidueIndexSelector(list_to_str(natro_residues))
natro_operation = rosetta.core.pack.task.operation.OperateOnResidueSubset(
rosetta.core.pack.task.operation.PreventRepackingRLT(),
natro_selector)
task_factory.push_back(natro_operation)
if extra_rotamers:
ers = rosetta.core.pack.task.operation.ExtraRotamersGeneric()
ers.ex1(True)
ers.ex2(True)
ers.extrachi_cutoff(18)
task_factory.push_back(ers)
if limit_aro_chi2:
lac = rosetta.protocols.task_operations.LimitAromaChi2Operation()
task_factory.push_back(lac)
if layered_design:
ld = rosetta.protocols.rosetta_scripts.XmlObjects.static_get_task_operation(
'''<LayerDesign name="layer_all" layer="core_boundary_surface_Nterm_Cterm" use_sidechain_neighbors="True">
<Nterm>
<all append="DEGHKNQRST" />
<all exclude="CAFILMPVWY" />
</Nterm>
<Cterm>
<all append="DEGHKNQRST" />
<all exclude="CAFILMPVWY" />
</Cterm>
</LayerDesign>''')
task_factory.push_back(ld)
return task_factory
def prepare_row(row: str) -> str:
prepared_row = list(map(lambda char: f"{char} ", f"|{row}|"))
return list_to_str(prepared_row)
def stringify(self, depth=0):
if self.tag == "[document]":
return self.children[0].stringify()
tabs = ""
for i in range(depth):
tabs += "\t"
if self.tag == "text":
if self.value is None:
return ""
string = self.value
for c in self.children:
if c.removed or c.is_empty_text():
continue
string += c.stringify(depth + 1)
regex_bl = ""
regex_br = ""
regex_symbol = ""
if self.optional:
regex_bl = "("
regex_br = ")"
regex_symbol = "?"
if self.repeating:
regex_bl = "("
regex_br = ")"
regex_symbol = "+"
string = f"{tabs}{regex_bl}{string}{regex_br}{regex_symbol}"
else:
attrs_list = ""
for attr_name in self.attrs:
if type(self.attrs[attr_name]) is list:
attr_string = list_to_str(self.attrs[attr_name])
else:
attr_string = self.attrs[attr_name]
attrs_list += f' {attr_name}="{attr_string}"'
opening_tag = f"<{self.tag}{attrs_list}>"
if self.tag != "br":
closing_tag = f"</{self.tag}>"
else:
closing_tag = ""
children_string = ""
regex_bl = ""
regex_br = ""
regex_symbol = ""
if self.optional:
regex_bl = "("
regex_br = ")"
regex_symbol = "?"
if self.repeating:
regex_bl = "("
regex_br = ")"
regex_symbol = "+"
if len(self.children) > 0:
if len(self.children) == 1 and self.children[
0].tag == "text" and not self.children[
0].is_empty_text():
string = f"{tabs}{regex_bl}{opening_tag}{self.children[0].stringify()}{closing_tag}{regex_br}{regex_symbol}"
else:
for c in self.children:
if c.removed or c.is_empty_text():
continue
children_string += c.stringify(depth + 1) + "\n"
string = f"{tabs}{regex_bl}{opening_tag}\n{children_string}{tabs}{closing_tag}{regex_br}{regex_symbol}"
else:
string = f"{tabs}{regex_bl}{opening_tag}{closing_tag}{regex_br}{regex_symbol}"
return string
def cmd_say(self, msg):
"""Say what the user told us to say
msg - list of strings
"""
return self.send_msg(utils.list_to_str(msg))
self.p = p
def encrypt(self, m):
res = []
for i in range(len(m)):
j = i % len(self.k)
res.append((m[i] + self.k[j]) % self.p)
return res
def decrypt(self, c):
res = []
for i in range(len(c)):
j = i % len(self.k)
res.append((c[i] - self.k[j]) % self.p)
return res
if __name__ == "__main__":
enc = str_to_list("wegvrjepqnwezbdzqetgzbigvpta")
affine = Affine(exeu(5, 26), 13)
res = affine.decrypt(enc)
res = list_to_str(res)
print(res)
key = intstr_to_list("20170317")
vigenere = Vigenere(key, 10)
m = intstr_to_list("314159265358979323846264")
res = vigenere.encrypt(m)
res = list_to_intstr(res)
print(res)
latin_chars.add("œ")
# remove some spurious characters
latin_chars.remove("<")
latin_chars.remove(">")
latin_chars.remove(",")
print(latin_chars)
char_embeddings = {}
char2index = {char: i for i, char in enumerate(latin_chars)}
for char, one_hot_vector in zip(
char2index.keys(),
keras.utils.to_categorical(list(char2index.values()),
num_classes=len(char2index))):
char_embeddings[char] = one_hot_vector
assert len(char_embeddings) == len(latin_chars), "Size of embedding and character set mismatch, expected {}, got {}" \
.format(len(latin_chars), len(char_embeddings))
print("Extracted char embeddings")
latin_path = Path("../data/alphabets/latin.csv")
latin_path.touch()
out = latin_path.open('w', encoding='utf-16')
# empty header
out.write("\n")
for char, embedding in char_embeddings.items():
s = char + "," + list_to_str(embedding) + "\n"
out.write(s)
out.close()
