def setup(self):
if RUN_NETWORK_TESTS:
time.sleep(1) # Reduce test interdependence
self.got_response = False
self.got_timeout = False
self.got_error = False
self.found_nodes = False
self.got_routing_response = False
self.got_routing_error = False
self.got_routing_timeout = False
self.got_routing_nodes_found = False
self.querier_mock = querier.QuerierMock(tc.CLIENT_ID)
self.r = minitwisted.ThreadedReactor(task_interval=.01)
self.rpc_m = rpc_manager.RPCManager(self.r,
tc.CLIENT_ADDR[1])
self.querier = querier.Querier(self.rpc_m,
tc.CLIENT_NODE)
self.querier_routing = querier.Querier(self.rpc_m,
tc.CLIENT_NODE)
self.querier_routing.set_on_response_received_callback(
self.on_routing_response)
self.querier_routing.set_on_error_received_callback(
self.on_routing_error)
self.querier_routing.set_on_timeout_callback(
self.on_routing_timeout)
self.querier_routing.set_on_nodes_found_callback(
self.on_routing_nodes_found)
self.r.start()
def setup(self):
import rpc_manager
global time
time = minitwisted.time = querier.time = MockTime()
self.got_response = False
self.got_timeout = False
self.got_error = False
self.querier = querier.Querier(tc.CLIENT_ID)
self.r = minitwisted.ThreadedReactor(task_interval=.01)
self.rpc_m = rpc_manager.RPCManager(self.r)
self.querier = querier.Querier(tc.CLIENT_NODE)
self.r.start()
def setup(self):
self.got_peers = None
querier_ = querier.Querier(tc.CLIENT_ID)
routing_m = RoutingManagerMock()
self.bootstrap_nodes = routing_m.get_closest_rnodes(tc.INFO_HASH_ZERO)
self.lm = lookup_manager.LookupManager(tc.CLIENT_ID, querier_,
routing_m, 2)
self.lookup = self.lm.get_peers(tc.INFO_HASH, self._on_got_peers,
tc.BT_PORT)
def setup(self):
time.mock_mode()
self.controller = controller.Controller(PYMDHT_VERSION, tc.CLIENT_NODE,
'test_logs/state.dat',
routing_m_mod, lookup_m_mod,
exp_m_mod, None)
self.my_id = self.controller._my_id
self.querier2 = querier.Querier() #self.my_id)
self.servers_msg_f = message.MsgFactory(PYMDHT_VERSION, tc.SERVER_ID)
# Authors: Thierry Moudiki
#
# License: BSD 3
import pandas as pd
import querier as qr
# Import data -----
url = ('https://raw.github.com/pandas-dev'
'/pandas/master/pandas/tests/data/tips.csv')
df = pd.read_csv(url)
# Example 1 -----
qrobj = qr.Querier(df=df)
df1 = qrobj\
.select(req="tip, sex, smoker, time")\
.filtr(req="smoker == 'No'")\
.summarize(req="sum(tip), sex, time", group_by="sex, time")\
.df
print(df1)
# Example 2 -----
df2 = qr.Querier(df)\
.select(req='tip, sex, day,size')\
.filtr(req="(day == 'Sun') | (day == 'Sat')")\
.summarize(req="avg(tip), sex, day", group_by="sex, day")\
def corrtarget_encoder(df, target, rho=0.4, verbose=1, seed=123):
""" Encode non-numerical columns using correlations.
Parameters
----------
df: a data frame
a data frame
target: str
target column a.k.a response
rho: float
correlation between pseudo target (used for averaging) and target
verbose: int
currently 0 = nothing printed; 1 = progress bar printed
seed: int
reproducibility seed
Returns
--------
a tuple: numerical data frame and achieved correlation
"""
target_ = df[target].values
target_mean = target_.mean()
target_std = target_.std()
n_target = len(target_)
C = np.eye(2)
C[0, 1] = rho
C[1, 0] = rho
C_ = np.linalg.cholesky(C).T
np.random.seed(seed)
temp = np.vstack((target_, np.random.normal(size=n_target))).T
df_ = pickle.loads(pickle.dumps(df, -1))
df_["pseudo_target"] = target_mean + np.dot(temp, C_)[:, 1] * target_std
covariates_names = df.columns.values[df.columns.values != target].tolist()
X = qr.select(df, ", ".join(covariates_names))
X_dtypes = X.dtypes
X_numeric = pickle.loads(pickle.dumps(df, -1))
col_iterator = covariates_names if verbose == 0 else tqdm(covariates_names)
for col in col_iterator:
if X_dtypes[col] == np.object: # something like a character string
X_temp = qr.summarize(df_,
req=col + ", avg(pseudo_target)",
group_by=col)
levels = np.unique(qr.select(X, col).values)
for l in levels:
qrobj = qr.Querier(X_temp)
val = np.float(
qrobj.filtr(col + '== "' + l +
'"').select("avg_pseudo_target").df.values)
X_numeric = qr.setwhere(X_numeric,
col=col,
val=l,
replace=val,
copy=False)
else: # a numeric column
X_numeric[col] = X[col]
return (
X_numeric,
np.corrcoef(df[target].values, df_["pseudo_target"].values)[0, 1],
)
#
# License: BSD 3
import pandas as pd
import querier as qr
import sqlite3
import sys
# data -----
url = ('https://raw.github.com/pandas-dev'
'/pandas/master/pandas/tests/data/tips.csv')
# Example 1 - Import from csv -----
qrobj1 = qr.Querier(source=url)
df1 = qrobj1\
.select(req="tip, sex, smoker, time")\
.filtr(req="smoker == 'No'")\
.summarize(req="sum(tip), sex, time", group_by="sex, time")\
.df
print(df1)
# Example 2 - Import from sqlite3 -----
# an sqlite3 database connexion
con = sqlite3.connect('people.db')
with con:
