pynut_tools - Laurent Tupin

It provides various functions to simplify the users life.

Installation

You can install the package from PyPI:

python -m pip install pynut-tools

The package is supported on Python 3.7 and above.

How to use

You can call a function as this example:

$ ----------------------------------------------------
>>> from pyNutTools import nutDate
>>> nutDate.fDte_Today()

This is the libraries I am using with the package

$ ----------------------------------------------------
>>> pandas==1.1.3

Documentation

Temporary documentation for nutDate :

from pyNutTools import nutDate as dat

dte_date = dat.fDte_formatToDate(dte_date, str_dateFormat = '%d/%m/%Y')
""" fDte_formatToDate makes sure you will have a varable with a date format
    The first Argument is the Variable (date), and the format of the string if it is a sting
    It allows you to avoid testing the type of the variable and get your get Date anyhow"""

int_dateDiff = dat.fInt_dateDifference(dte_date1, dte_date2)
""" fInt_dateDifference give you the difference in days between 2 dates"""

Date2 = dat.fDte_convertExcelInteger(Date)
""" fDte_convertExcelInteger takes an integer as input, 
    This is the integer you can find in Excel when it is a date 
    And return the associated date  """

Temporary documentation for nutDataframe :

from pyNutTools import nutDataframe as dframe

bl_isempty = dframe.fBl_isDataframeEmpty(df_simple)
""" Test if a Dataframe is empty"""

df_simple = dframe.fDf_createSimpleDataframe()
""" Create a simple dataframe to make test"""

bl_compare, df_compare = dframe.fBl_compareDfCol({'df': df_1, 'colJoin': 'colJoin','colToCompare':'data'},
                                                  {'df': df_2,'colJoin': 'colJoin','colToCompare':'data'})
""" compare 2 dataframe one a numeric column by joining the df and returning the difference """

df_1['DataRounded'] = df_1['DataToBeRounded'].apply(lambda x: dframe.round_down(x))
""" Use the Math Function floor() - Able to add a decimals like in Excel
    floor() rounds down. int() truncates. 
    The difference is clear when you use negative numbers
    math.floor(-3.5)    -4
    int(-3.5)           -3"""

df_2['DataRounded'] = df_2['DataToBeRounded'].apply(lambda x: dframe.round_up(x))
""" Use the Math Function ceil() - Able to add a decimals like in Excel"""

df_data = dframe.fDf_readCsv_enhanced(path, bl_header = None, str_sep = '|', l_names = range(33))
""" Use the pandas method read_csv
    but resolving Parse Error and will try again after displaying a message 
    Also resolving UnicodeDecodeError by detecting the encoding and trying again accordingly """
 
df2 = dframe.fDf_removeDoublons(df1)
""" Remove all rows that are exactly the same"""
 
df2 = dframe.fDf_DropRowsIfNa_resetIndex(df1, l_colToDropNA = ['col1'])
""" Drop the rows where all defined columns will be Nan
    And reset the index"""

df2 = dframe.dDf_fillNaColumn(df1, 'col2', 'col1')
""" Replace Nan in a column by the value in another column or a Constant """

df2 = dframe.fDf_fillColUnderCondition(df1, 'NameColApply', df1['data'], 'NameColC', 'YES', bl_except = False)
''' Transform DF with condition
    ValueToApply can be a value or a lambda function'''   

Temporary documentation for nutOther :

from pyNutTools import nutOther as oth

1. Decorators

@oth.dec_singletonsClass
class CLASS_TO_DECORATE():
''' Singeltons decorators: always use the first instance 
    Example: connection to database, FTP (keep the same connection for performance and possibly Access issue)
    '''    

@oth.dec_getTimePerf(int_secondesLimitDisplay = 2)
def function_TO_DECORATE(*args, **kwarks):
''' Time Performance Decorators on a function
    You can calculate and compare Performance on any function just by decorating it
    It will show nothing if the performance is better than a specific threshold you will defined
    '''   

@oth.dec_stopProcessTimeOut(int_secondesLimit = 10, returnIfTimeOut = False)
def function_TO_DECORATE(*args, **kwarks):
''' This decorators allow to stop a process if it is too long
    For example, testing a folder existence might be very very long...'''

END