jmquintana79

## custom function
def custom_function_example(x,y):
  return x + y
## apply rolling on a custom function
def apply_rolling_function(df, window_size, func, column1, column2, verbose = False):
    # validate arguments
    assert column1 in df.columns.tolist()
    assert column2 in df.columns.tolist()
    # initialize
    result = []

jmquintana79

## mapplot creation of a z variable according to x/y variables, all of them, in a df
def mapplot(df:pd.DataFrame, c_x:str, c_y:str, c_z:str, title:str = '', c_map:str = "rainbow"):
    # validate arguments
    assert c_x in df.columns.tolist()
    assert c_y in df.columns.tolist()
    assert c_z in df.columns.tolist() 
    # initialize 
    import matplotlib.pyplot as plt
    # collect data
    x = df[c_x].values 

jmquintana79

# multiple pandas df tables to one excel file on multiple sheets
with pd.ExcelWriter(path_output, engine='xlsxwriter') as writer:
  df1.to_excel(writer, sheet_name='sheet1')
  df2.to_excel(writer, sheet_name='sheet2')
  df3.to_excel(writer, sheet_name='sheet3')

jmquintana79

import pandas as pd
from sklearn.preprocessing import FunctionTransformer
from sklearn.pipeline import Pipeline
# example
from sklearn.linear_model import LogisticRegression

# X, y

def get_dummies_size(df):
    return pd.get_dummies(df, columns=['size'])

jmquintana79

import pandas as pd
import numpy as np
from sklearn.pipeline import Pipeline
from sklearn.compose import ColumnTransformer
# example models and preprocessors
from sklearn.preprocessing import StandardScaler, OneHotEncoder
from sklearn.impute import SimpleImputer
from sklearn.linear_model import LogisticRegression

# X, y

jmquintana79

Reference: https://hydra.cc/docs/patterns/configuring_experiments/

config.ymal

defaults:
  - db: mysql
  - server: apache

db/mysql.ymal

jmquintana79

class Credentials():

    def __init__(self):
        self.user = "user"
        self.password = "password"


class Service(Credentials):
    def __init__(self):
        super().__init__()

jmquintana79

import pandas as pd

## unstack a timeseries target variable according to a categorical reference column 
def unstack_ts_according_to_reference(df:pd.DataFrame, c_dt:str, c_cat_reference:str, c_target_variable:str)->pd.DataFrame:
    """
    Unstack a timeseries target variable according to a categorical reference column.
    df -- Dataframe to be processed.
    c_dt -- Temporal column.
    c_cat_reference -- Categorical column to be used as reference to stack the target variable.
    c_target_variable -- Num / Cat column to be stacked.

jmquintana79

from scipy.stats import linregress
# estimate linear regression y = Ax + B
A, B, r_value, p_value, std_err = linregress(x, y)

jmquintana79

# original column
In [15]: df["timedelta_column"]
Out[15]: 
0    1 days 00:00:00
1    3 days 02:00:00
2    5 days 04:00:00
3    7 days 06:00:00
4    9 days 08:00:00
5   11 days 10:00:00
dtype: timedelta64[ns]