Introduction to conjurer"

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Overview

Background & Motivation

Data science applications need data to prototype and demonstrate to potential clients. For such purposes, using production data is a possibility. However, it is not always feasible due to legal and/or ethical considerations[@SynDataNeed]. This resulted in a need for generating synthetic data. This need is the key motivator for the package conjurer. This package is under constant development and the author would update the documentation regularly at FOYI and uncovr

Need for package conjurer

Data across multiple domains are known to exhibit some form of seasonality, cyclicality and trend. Although there are synthetic data generation packages currently available, they focus primarily on synthetic versions of microdata containing confidential information or for machine learning purposes. There is a need for a more generic synthetic data generation package that helps for multiple purposes such as forecasting, customer segmentation, insight generation etc. This package conjurer helps in generating such synthetic data.

Steps to build synthetic data

Let us consider an example of generating transactional data for a retail store. The following steps will help in building such data.

Installation

Install conjurer package by using the following code. Since the package uses base R functions, it does not have any dependencies.

install.packages("conjurer")

Build customers

A customer is identified by a unique customer identifier(ID). A customer ID is alphanumeric with prefix "cust" followed by a numeric. This numeric ranges from 1 and extend to the number of customers provided as the argument within the function. For example, if there are 100 customers, then the customer ID will range from cust001 to cust100. This ensures that the customer ID is always of the same length. Let us build a group of customer IDs using the following code. For simplicity, let us assume that there are 100 customers. customer ID is built using the function buildCust. This function takes one argument "numOfCust" that specifies the number of customer IDs to be built.

library(conjurer)
customers <- buildCust(numOfCust =  100)
print(head(customers))

Build customer names

A list of customer names for the 100 customer IDs can be generated in the following way.

custNames <- as.data.frame(buildNames(numOfNames = 100, minLength = 5, maxLength = 7))

#set column heading
colnames(custNames) <- c("customerName")
print(head(custNames))

Assign customer name to customer ID

Let us assign customer names to customer IDs. This is a random one to one mapping using the following code.

customer2name <- cbind(customers, custNames)
#set column heading
print(head(customer2name))

Build customer age

A list of customer ages for the 100 customer IDs can be generated in the following way.

custAge <- as.data.frame(round(buildNum(n = 10, st = 23, en = 80, disp = 0.5, outliers = 1)))

#set column heading
colnames(custAge) <- c("customerAge")
print(head(custAge))

Assign customer age to customer ID

Let us assign customer ages to customer IDs. This is a random one to one mapping using the following code.

customer2age <- cbind(customers, custAge)
#set column heading
print(head(customer2age))

Build customer phone number

A list of customer phone numbers for the 100 customer IDs can be generated in the following way.

parts <- list(c("+91","+44","+64"), c("("), c(491,324,211), c(")"), c(7821:8324))
probs <- list(c(0.25,0.25,0.50), c(1), c(0.30,0.60,0.10), c(1), c())
custPhoneNumbers <- as.data.frame(buildPattern(n=100,parts = parts, probs = probs))
head(custPhoneNumbers)

#set column heading
colnames(custPhoneNumbers) <- c("customerPhone")
print(head(custPhoneNumbers))

Assign customer phone number to customer ID

Let us assign customer ages to customer IDs. This is a random one to one mapping using the following code.

customer2phone <- cbind(customers, custPhoneNumbers)
#set column heading
print(head(customer2phone))

Build products

The next step is building some products. A product is identified by a product ID. Similar to a customer ID, a product ID is also an alphanumeric with prefix "sku" which signifies a stock keeping unit. This prefix is followed by a numeric ranging from 1 and extending to the number of products provided as the argument within the function. For example, if there are 10 products, then the product ID will range from sku01 to sku10. This ensures that the product ID is always of the same length. Besides product ID, the product price range must be specified. Let us build a group of products using the following code. For simplicity, let us assume that there are 10 products and the price range for them is from 5 dollars to 50 dollars. Products are built using the function buildProd. This function takes 3 arguments as given below.

products <- buildProd(numOfProd = 10, minPrice = 5, maxPrice = 50)
print(head(products))

Build product hierarchy

The products belong to various categories. Let's start to build the product hierarchy. The 10 products belong to 2 categories namely Food and Non-Food. These categories are further classifed into 4 different sub-categories namely Beverages, Dairy, Sanitary and Household.

productHierarchy <- buildHierarchy(type = "equalSplit", splits = 2, numOfLevels = 2)
print(productHierarchy)

As you can see, the product hierarchy generated has default names for levels and elements. To make it more meaningful, it can be modified as follows.

#Rename the dataframe
names(productHierarchy) <- c("category", "subcategory")

#Replace category with Food and Non-Food
productHierarchy$category <- gsub("Level_1_element_1", "Food", productHierarchy$category)
productHierarchy$category <- gsub("Level_1_element_2", "Non-Food", productHierarchy$category)

#Replace subCategories
productHierarchy$subcategory <- gsub("Level_2_element_1", "Beverages", productHierarchy$subcategory)
productHierarchy$subcategory <- gsub("Level_2_element_3", "Dairy", productHierarchy$subcategory)
productHierarchy$subcategory <- gsub("Level_2_element_2", "Sanitary", productHierarchy$subcategory)
productHierarchy$subcategory <- gsub("Level_2_element_4", "Household", productHierarchy$subcategory)

#Inspect the data to confirm the results 
productHierarchy <- productHierarchy[order(productHierarchy$category),]
print(productHierarchy)

Build transactions

Now that a group of customer IDs and Products are built, the next step is to build transactions. Transactions are built using the function genTrans. This function takes 5 arguments. The details of them are as follows.

Let us build transactions using the following code

transactions <- genTrans(cycles = "y", spike = 12, outliers = 1, transactions = 10000)

Visualize generated transactions by using

TxnAggregated <- aggregate(transactions$transactionID, by = list(transactions$dayNum), length)
plot(TxnAggregated, type = "l", ann = FALSE)

Build final data

Bringing customers, products and transactions together is the final step of generating synthetic data. This process entails 3 steps as given below.

Allocate customers to transactions

The allocation of transactions is achieved with the help of buildPareto function. This function takes 3 arguments as detailed below.

Let us now allocate transactions to customers first by using the following code.

customer2transaction <- buildPareto(customers, transactions$transactionID, pareto = c(80,20))

Assign readable names to the output by using the following code.

names(customer2transaction) <- c('transactionID', 'customer')

#inspect the output
print(head(customer2transaction))

Allocate products to product hierarchy

Allocate the products to the product hierarchy. This can be achieved as follows.

#First step is to ensure that the product hierarchy data frame has the same number of rows as number of products.
category <- productHierarchy$category
subcategory <- productHierarchy$subcategory
productHierarchy <- as.data.frame(cbind(category,subcategory,1:nrow(products)))

#Randomly assign the product hierarchy to the products. Ensure that the additional unused variable towards the end is dropped.
products <- cbind(products, productHierarchy[,c("category","subcategory")])
#inspect the output
print(head(products))

Allocate products to transactions

Now, using similar step as mentioned above, allocate transactions to products using following code.

product2transaction <- buildPareto(products$SKU,transactions$transactionID,pareto = c(70,30))
names(product2transaction) <- c('transactionID', 'SKU')

#inspect the output
print(head(product2transaction))

Combine customers and transactions data

The following code brings together transactions, products and customers into one dataframe.

df1 <- merge(x = customer2transaction, y = product2transaction, by = "transactionID")

df2 <- merge(x = df1, y = transactions, by = "transactionID", all.x = TRUE)

#inspect the output
print(head(df2))

Final data

We can add additional data such as customer name, product price using the code below.

df3 <- merge(x = df2, y = customer2name, by.x = "customer", by.y = "customers", all.x = TRUE)
df4 <- merge(x = df3, y = customer2age, by.x = "customer", by.y = "customers", all.x = TRUE)
df5 <- merge(x = df4, y = customer2phone, by.x = "customer", by.y = "customers", all.x = TRUE)
df6 <- merge(x = df5, y = products, by = "SKU", all.x = TRUE)
dfFinal <- df6[,c("dayNum", "mthNum", "customer", "customerName", "customerAge", "customerPhone", "transactionID", "SKU", "Price", "category","subcategory")]


#inspect the output
print(head(dfFinal))

Thus, we have the final data set with transactions, customers and products.

Interpret the results

The column names of the final data frame can be interpreted as follows.

Let us visualize the results to understand the data distribution.

Below is a view of the sum of transactions by each day.

aggregatedDataDay <- aggregate(dfFinal$transactionID, by = list(dfFinal$dayNum), length)
plot(aggregatedDataDay, type = "l", ann = FALSE)

Below is a view of the sum of transactions by each month.

aggregatedDataMth <- aggregate(dfFinal$transactionID, by = list(dfFinal$mthNum), length)
aggregatedDataMthSorted <- aggregatedDataMth[order(aggregatedDataMth$Group.1),]
plot(aggregatedDataMthSorted, ann = FALSE)

r '## References {-}'



Try the conjurer package in your browser

Any scripts or data that you put into this service are public.

conjurer documentation built on May 1, 2022, 9:05 a.m.