README.Rmd

---
title: "easyRFM - An easy way to RFM analysis by R"
author: "Koji MAKIYAMA"
output: 
  html_document: 
    keep_md: true
---

```{r, echo=FALSE}
knitr::opts_chunk$set(echo=TRUE, warning=FALSE, message=FALSE)
```

## Overview

About RFM analysis:

- [RFM (customer value) - Wikipedia](http://en.wikipedia.org/wiki/RFM_%28customer_value%29)

> RFM is a method used for analyzing customer value. It is commonly used in database marketing and direct marketing and has received particular attention in retail and professional services industries.
> 
> RFM stands for
>
> - Recency - How recently did the customer purchase?
> - Frequency - How often do they purchase?
> - Monetary Value - How much do they spend?

First, ready transaction data like below:

```{r, echo=FALSE}
library(easyRFM)
data <- rfm_generate_data(seed=123)
```

```{r}
head(data)
```

The "id" means customer ID, the "payment" means a payment for purchase and the "date" means a purchase date.

Then you can execute RFM analysis by a simple command:

```{r}
result <- rfm_auto(data)
```

The result contains three elements and one function.

`result$rfm` is which class each customer was assigned.

```{r}
head(result$rfm)
```

`result$breaks` is the breaks for each classes.

```{r}
result$breaks
```

`result$classes` is the ranges for each classes.

```{r}
result$classes
```

`result$get_table()` is function which creates tables with slicing.

```{r}
result$get_table("RF", M_slice=4:5)
```

If you don't indicate slice, it uses all.

```{r}
result$get_table("RF")
```

## How to install

```{r, eval=FALSE}
install.packages("devtools") # if you have not installed "devtools" package
devtools::install_github("hoxo-m/easyRFM")
```

## Try it with sample data

easyRFM package provide `rfm_generate_data()` function to generate sample data for `rfm_auto()`:

```{r}
data <- rfm_generate_data()
head(data)
```

Try `rfm_auto()` and look over the result:

```{r}
result <- rfm_auto(data)
```

## How to input to rfm_auto()

If your data have different column names from default: "id", "payment" and "date", for example:

```{r, echo=FALSE}
data <- rfm_generate_data(seed=123)
colnames(data) <- c("customer_id", "payment", "purchase_date")
```

```{r}
head(data)
```

You can indicate the column names:

```{r}
result <- rfm_auto(data, id="customer_id", payment="payment", date="purchase_date")
```

If your data have different date format from default: "yyyy-mm-dd", for example:

```{r, echo=FALSE}
library(stringr)
data <- rfm_generate_data(seed=123)
data$date <- str_replace_all(data$date, "-", "/")
```

```{r}
head(data)
```

You can indicate date format:

```{r}
result <- rfm_auto(data, date_format = "%Y/%m/%d")
```

For more information for date_format, see [Date-time Conversion Functions to and from Character](http://stat.ethz.ch/R-manual/R-patched/library/base/html/strptime.html).

You can use datetime object(POSIXlt or POSIXct) instead of date, for example:

```{r, echo=FALSE}
library(stringr)
data <- rfm_generate_data(seed=123, date_type = "POSIXct")
data$date <- format(data$date, "%Y/%m/%d %H:%M:%S")
```

```{r}
head(data)
```

```{r}
result <- rfm_auto(data, date_format = "%Y/%m/%d %H:%M:%S")
```

## Application

```{r}
data <- rfm_generate_data(10000, begin="2014-10-01", end="2015-01-01", seed=123)
result <- rfm_auto(data, breaks=list(r=6, f=5, m=5))

result$get_table("RF", M_slice=4:5)

leaved_customers <- result$get_sliced_rfm(R_slice=1:2, F_slice=2:5, M_slice=4:5)
leaving_customers <- result$get_sliced_rfm(R_slice=3:4, F_slice=4:5, M_slice=4:5)
good_customers <- result$get_sliced_rfm(R_slice=5:6, F_slice=4:5, M_slice=4:5)
```