workforcerappi 0.2.0

No project description provided

Rappi's Workforce Team Data Analytics and Forecasting

This Python package provides functionality for data analysis and forecasting for Rappi's Workforce Management (WFM) system.

Installation

You can install the package using pip:

pip install workforcerappi


# #The package includes functions for processing and preparing data for analysis.

# importar librerias
from workforcerappi import data_processing as dp, modeling, order_distribution as od
import pandas as pd

# definir parametros
month="Agosto"
startdate = "2024-03-01"
FEATURES = ['dayofweek', 'quarter', 'month', 'year', 'dayofyear', 'dayofmonth', 'weekofyear']
year_month="2024-08"
TARGET = "SS" # or "BR"
SERVICE_USER_TYPE="Users"
SERVICE="Customer Live Orders"
REGION="SS" # or "BR"
concurrencia = 2  # concurrencia del servicio
sla = 0.8  # nivel de servicio
meta_aht = 10  
date_range=['2023-01-01', '2024-05-31']  

# Read orders from Snowflake
df_ordered = dp.read_and_sort_orders(date_range)

#Read Special days from Snowflake and Preprocess it
special_days=dp.preprocess_special_dates(date_range)
df_filtered = dp.filter_special_dates(df_ordered, special_days)

# Pivot Orders
orders = dp.pivot_orders(df_filtered,REGION)
data = orders.set_index("FECHA")

#Split data into train and test sets
train, test = dp.split_train_test_data(data, startdate)

# Add features like month, day of the week, week of the year, etc. from date
train = dp.create_features(train)
test = dp.create_features(test)

X_train = train[FEATURES]
y_train = train[TARGET]
X_test = test[FEATURES]
y_test = test[TARGET]

# Train the model
model = modeling.train_xgboost_model(X_train, y_train, X_test, y_test)

# Make Predictions for next months
predictions = modeling.make_predictions(model, X_test)
test['prediction'] = predictions
predictions = dp.create_future_predictions(model, FEATURES, '2024-08-01','2024-08-31', 'D')
month_to_predict = predictions

# Create Forecasted Orders Distribution
ordenes_financieras = od.create_ordenes_financieras(df_ordered, month_to_predict, ordenes_aprobadas=None, REGION=REGION , year_month=year_month)

# Create Orders Curve 
curva_ordenes = od.create_curva_ordenes(df_ordered,REGION)
# plot orders curve
reporting.plot_orders_curve(curva_ordenes)
# PxQ forecasted orders by orders curve Distribution
df3 = od.distribute_orders(ordenes_financieras, curva_ordenes)
print(df3.head())

# ajustar dias de festividades
# valor_a_ajustar = 1.30
# df3.loc[df3["FECHA"] == "2024-08-07", "CO"] *= valor_a_ajustar

# Read Inflow, AHT, CR from Snowflake Database
inflow = dp.read_cr_aht(date_range)

# Distribute inflow by interval
result=od.distribute_inflow_intraday(inflow, df3, SERVICE_USER_TYPE, SERVICE, REGION)

# Distribute aht by interval
result2 = od.distribute_aht_intraday(inflow, df3, SERVICE_USER_TYPE, SERVICE,REGION, result, meta_aht)

# Calculate Headcount based in inflow,aht, sla and concurrency 
required_headcount_df = od.hc(result,result2,sla, concurrencia)
print(required_headcount_df.head(3))

# Calculate requered hours based in headcount 
od.required_hours(required_headcount_df)
print(required_hours.head(3))

print("Done!")


# Reporting Module:

# Time series:
reporting.plot_time_series(df_ordered, 'FECHA','ORDERS')

#Orders Curve 
reporting.plot_general_orders_curve(df_ordered, 'INTERVALO', 'ORDERS')

#Proportion of Orders by Country
reporting.plot_proportion_of_orders_by_country(df_ordered, 'ORDERS', 'COUNTRY')

# Total Numbers of orders per country
reporting.plot_total_orders_per_country(df_ordered, 'COUNTRY','ORDERS')

# Plot raw data vs predictions on test data
reporting.plot_raw_data_vs_prediction(data,X_test,test,model)

# Plot raw data vs forecasted
reporting.plot_raw_data_vs_forecasted(data,month_to_predict)

License

This project is licensed under a private license by Rappi Inc. Unauthorized copying, distribution, or modification of this project, via any medium, is strictly prohibited. For more details, please contact the legal department at Rappi.

Contact Information

For licensing inquiries, please reach out to:

• Email: [e-mail]

© 2024 Rappi Inc. Todos los derechos reservados