PK Oj4 4 logsensei/__init__.py"""Personal Formatting on Loguru"""
__version__ = "0.2.0"
__author__ = "Aditya Kelvianto Sidharta"
import logging
import os
import sys
from collections import Counter
import numpy as np
from loguru import logger as loguru_logger
from sklearn.metrics import (
accuracy_score,
explained_variance_score,
f1_score,
mean_absolute_error,
mean_squared_error,
precision_score,
r2_score,
recall_score,
roc_auc_score,
)
from torchsummary import summary
from logsensei.utils import _get_datetime
# pylint: disable=too-many-public-methods
class Logger:
"""
Setting up logger for the project. The log will be logged within the file as well
logger.setup_logger(script_name) must be called first before using the logger.
"""
def __init__(self, default_level=logging.INFO):
self.name = None
self.datetime = None
self.level = None
self.default_level = default_level
self.time_sys_format = "{time:YYYY-MM-DD HH:mm:ss}"
self.level_sys_format = "{level}"
self.function_sys_format = "{name}:{function}:{line}"
self.message_sys_format = "{message}"
self.sys_format = "{} | {} | {} | {}".format(
self.time_sys_format, self.level_sys_format, self.function_sys_format, self.message_sys_format
)
self.time_file_format = None
self.level_file_format = None
self.function_file_format = None
self.message_file_format = None
self.file_format = None
self.logger = loguru_logger
self.logger.remove(0)
self.logger.add(sys.stderr, format=self.sys_format, level=self.level)
self.logger.patch(lambda record: record.update(name="my_module"))
self.file_index = None
self.template = {}
def setup(self, name, logger_file, level=logging.DEBUG):
self.name = name
self.datetime = _get_datetime()
self.level = level
if self.file_index is not None:
self.logger.remove(self.file_index)
self.time_file_format = "{time:YYYY-MM-DD HH:mm:ss:SSS}"
self.level_file_format = "{level: <8}"
self.function_file_format = "{name: ^15}:{function: ^15}:{line: >3}"
self.message_file_format = "{message}"
self.file_format = "{} | {} | {} | {}".format(
self.time_file_format, self.level_file_format, self.function_file_format, self.message_file_format
)
self.file_index = self.logger.add(
os.path.join(logger_file, "{}_{}.log".format(self.name, self.datetime)),
format=self.file_format,
level=self.level,
)
def debug(self, msg):
self.logger.opt(depth=1).debug(msg)
def info(self, msg):
self.logger.opt(depth=1).info(msg)
def warning(self, msg):
self.logger.opt(depth=1).warning(msg)
def error(self, msg):
self.logger.opt(depth=1).error(msg)
def log(self, msg, level):
if level == logging.DEBUG:
self.debug(msg)
elif level == logging.INFO:
self.info(msg)
elif level == logging.WARNING:
self.warning(msg)
elif level == logging.ERROR:
self.error(msg)
def create(self, template_name, msg):
if template_name in self.template.keys():
self.logger.warning("Replacing the template message in {}".format(template_name))
self.template[template_name] = msg
def apply(self, template_name, *args):
self.log(self.template[template_name].format(args), self.default_level)
def df(self, df, df_name):
shape = df.shape
self.log("DataFrame {} shape : {}".format(df_name, shape), self.default_level)
self.log("DataFrame {} info:".format(df_name), self.default_level)
self.log(df.info(), self.default_level)
def array(self, array, array_name):
shape = array.shape
self.log("Array {} shape : {}".format(array_name, shape), self.default_level)
if array.ndim == 1:
n_values = len(array)
unique_values = set(array)
n_unique_values = len(set(array))
n_missing_values = np.sum(np.isnan(array))
self.log("Array {} unique values : {}".format(array_name, unique_values), self.default_level)
self.log("Array {} cardinality : {}".format(array_name, n_unique_values), self.default_level)
self.log(
"Array {} missing values : {} ({:.2f}%)".format(
array_name, n_missing_values, n_missing_values / n_values * 100.0
),
self.default_level,
)
if (array.dtype == float) or (array.dtype == int):
mean_value = np.nanmean(array)
std_value = np.nanstd(array)
max_value = np.nanmax(array)
min_value = np.nanmin(array)
median_value = np.nanmedian(array)
perc_25_value = np.nanpercentile(array, 25)
perc_75_value = np.nanpercentile(array, 75)
self.log(
"Array {} info : MEAN={} | STD={} | MIN={} | 25TH={} | MEDIAN={} | 75TH={} | MAX={}".format(
array_name,
mean_value,
std_value,
min_value,
perc_25_value,
median_value,
perc_75_value,
max_value,
),
self.default_level,
)
else:
most_common = Counter(array).most_common(5)
n_most_common = len(most_common)
self.log(
"Array {} top {} values : ".format(array_name, n_most_common)
+ " | ".join(["{} - {}({:.2f}%)".format(x[0], x[1], x[1] / n_values * 100.0) for x in most_common]),
self.default_level,
)
def compare(self, array_1, array_2, array_1_name, array_2_name):
array_1_shape = array_1.shape
array_2_shape = array_2.shape
array_1_ndim = array_1.ndim
array_2_ndim = array_2.ndim
self.log(
"Compare {} vs {} - shape : {} vs {}".format(array_1_name, array_2_name, array_1_shape, array_2_shape),
self.default_level,
)
if (array_1_ndim == 1) and (array_2_ndim == 1):
cardinality = len(set(array_1).union(set(array_2)))
array_1_unique = set(array_1)
array_2_unique = set(array_2)
array_intersection = array_1_unique.intersection(array_2_unique)
array_1_outer = array_1_unique - array_intersection
array_2_outer = array_2_unique - array_intersection
self.log("Compare {} vs {} - cardinality :".format(array_1_name, array_2_name), self.default_level)
self.log(
"Intersection {} and {} : {} ({:.2f}%)".format(
array_1_name, array_2_name, len(array_intersection), len(array_intersection) / cardinality
),
self.default_level,
)
self.log(
"Unique Values in {} but not in {} : {} ({:.2f}%)".format(
array_1_name, array_2_name, len(array_1_outer), len(array_1_outer) / cardinality
),
self.default_level,
)
self.log(
"Unique Values in {} but not in {} : {} ({:.2f}%)".format(
array_2_name, array_1_name, len(array_2_outer), len(array_2_outer) / cardinality
),
self.default_level,
)
def dict(self, dictionary, dictionary_name):
n_values = len(dictionary)
self.log("Dictionary {} length : {}".format(dictionary_name, n_values), self.default_level)
self.log("Dictionary {}".format(dictionary_name), self.default_level)
for key, value in dictionary.items():
self.log("{} - {}".format(key, value), self.default_level)
def list(self, input_list, input_list_name):
n_values = len(input_list)
self.log("List {} length : {}".format(input_list_name, n_values), self.default_level)
self.log("List {} : {}".format(input_list_name, input_list), self.default_level)
def set(self, input_set, input_set_name):
n_values = len(input_set)
self.log("Set {} length : {}".format(input_set_name, n_values), self.default_level)
self.log("Set {} : {}".format(input_set_name, input_set), self.default_level)
def savepath(self, file_to_save, save_path):
self.log("Saving {} to path : {}".format(file_to_save, save_path), self.default_level)
def loadpath(self, file_to_load, load_path):
self.log("Loading {} from path : {}".format(file_to_load, load_path), self.default_level)
def scikit(self, model, model_name):
self.log("Model {} type : {}".format(model_name, type(model).__name__), self.default_level)
self.dict(model.get_params(), "Parameters of scikit-learn model {}".format(model_name))
def xgboost(self, model, model_name):
raise NotImplementedError
def lightgbm(self, model, model_name):
raise NotImplementedError
def pytorch_tensor(self, tensor, tensor_name):
shape = tuple(tensor.shape)
if tensor.is_cuda:
array = tensor.cpu().numpy()
else:
array = tensor.numpy()
max_value = np.max(array)
min_value = np.min(array)
mean_value = np.mean(array)
median_value = np.median(array)
self.log("Tensor {} shape : {}".format(tensor_name, shape), self.default_level)
self.log(
"Tensor {} info : MEAN={} | MIN={} | MEDIAN={} | MAX={}".format(
tensor_name, mean_value, min_value, median_value, max_value
),
self.default_level,
)
def tensorflow_tensor(self, tensor, tensor_name):
shape = tuple(tensor.shape)
array = tensor.numpy()
max_value = np.max(array)
min_value = np.min(array)
mean_value = np.mean(array)
median_value = np.median(array)
self.log("Tensor {} shape : {}".format(tensor_name, shape), self.default_level)
self.log(
"Tensor {} info : MEAN={} | MIN={} | MEDIAN={} | MAX={}".format(
tensor_name, mean_value, min_value, median_value, max_value
),
self.default_level,
)
def pytorch_model(self, model, model_name, input_shape):
self.log("PyTorch Model {} Summary :".format(model_name), self.default_level)
self.log(summary(model, input_shape), self.default_level)
def tensorflow_model(self, model, model_name):
self.log("TensorFlow Model {} Summary :".format(model_name), self.default_level)
self.log(model.summary(), self.default_level)
def regression(self, true_array, predict_array, array_name):
self.log("{} Regression Score".format(array_name), self.default_level)
self.log("=" * 20, self.default_level)
self.log("Mean Absolute Error : {}".format(mean_absolute_error(true_array, predict_array)), self.default_level)
self.log("Mean Squared Error : {}".format(mean_squared_error(true_array, predict_array)), self.default_level)
self.log("R2 Score : {}".format(r2_score(true_array, predict_array)), self.default_level)
self.log(
"Explained Variance Score : {}".format(explained_variance_score(true_array, predict_array)),
self.default_level,
)
def classification(self, true_array, predict_array, array_name):
self.log("{} Classification Score".format(array_name), self.default_level)
self.log("=" * 20, self.default_level)
self.log("Accuracy Score : {}".format(accuracy_score(true_array, predict_array)), self.default_level)
self.log("Precision Score : {}".format(precision_score(true_array, predict_array)), self.default_level)
self.log("Recall Score : {}".format(recall_score(true_array, predict_array)), self.default_level)
self.log("F1 Score : {}".format(f1_score(true_array, predict_array)), self.default_level)
self.log("ROC AUC Score : {}".format(roc_auc_score(true_array, predict_array)), self.default_level)
def multiclass(self, true_array, predict_array, array_name):
self.log("{} Classification Score".format(array_name), self.default_level)
self.log("=" * 20, self.default_level)
self.log("Accuracy Score : {}".format(accuracy_score(true_array, predict_array)), self.default_level)
self.log(
"Precision Score : {}".format(precision_score(true_array, predict_array, average="micro")),
self.default_level,
)
self.log(
"Recall Score : {}".format(recall_score(true_array, predict_array, average="micro")), self.default_level
)
self.log("F1 Score : {}".format(f1_score(true_array, predict_array, average="micro")), self.default_level)
self.log(
"ROC AUC Score : {}".format(roc_auc_score(true_array, predict_array, average="micro")), self.default_level
)
logger = Logger()
PK sO.-d d logsensei/utils.pyimport datetime
def _get_datetime():
return datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
PK sO=%: : ! logsensei-0.2.0.dist-info/LICENSEMIT License
Copyright (c) 2019 Aditya Kelvianto Sidharta
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
PK !HMuS a logsensei-0.2.0.dist-info/WHEEL HM
K-*ϳR03rOK-J,/RH,szd&Y)r$[)T&UD" PK !HR\ " logsensei-0.2.0.dist-info/METADATA]Pn0 }Wy4jK˳nF8A/ě}.>ٓ`iXwBKzo"HUWNb5LwÓEvޤԳ
hSuQ.}卷EٲXsa,ṽj[ >+U&|pq7UE5Y6Ƚ9O=# KaD7o7ZO@1{(>hl% [r.a\CKPK !H1C) logsensei-0.2.0.dist-info/RECORD}MB@ {h" F+ \5bV}^~b{1GBMhk}4[J<ǨvmL:ډYI/V3v8.!*},V\lҊ!4}PK Oj4 4 logsensei/__init__.pyPK sO.-d d 4 logsensei/utils.pyPK sO=%: : ! Y5 logsensei-0.2.0.dist-info/LICENSEPK !HMuS a 9 logsensei-0.2.0.dist-info/WHEELPK !HR\ " b: logsensei-0.2.0.dist-info/METADATAPK !H1C) ; logsensei-0.2.0.dist-info/RECORDPK <