slowly-changing-dimension.Rmd

---
title: "Slowly Changing Dimension methodology"
output: rmarkdown::html_vignette
bibliography: "references.bib"
vignette: >
  %\VignetteIndexEntry{Slowly Changing Dimension methodology}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---
```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

```{r library, eval = requireNamespace("tidyverse")}
library(tidyverse)
```

```{r setup, include = FALSE, eval = requireNamespace("tidyverse")}

forecasts_full <- tribble(
  ~City, ~Forecast, ~ForecastDate,
  "New York",    20, "2023-09-28",
  "Los Angeles", 23, "2023-09-28",
  "Seattle",     16, "2023-09-28",
  "Houston",     34, "2023-09-28",
  "New York",    18, "2023-09-29",
  "Los Angeles", 25, "2023-09-29",
  "Seattle",     17, "2023-09-29",
  "Houston",     34, "2023-09-29"
) %>%
  mutate(across("ForecastDate", as.Date))

forecasts_scd <- forecasts_full %>%
  mutate(ForecastFrom = ForecastDate,
         ForecastUntil = case_when(
           City == "Houston" ~ NA,
           ForecastDate == "2023-09-29" ~ NA,
           TRUE ~ "2023-09-29"
         )) %>%
  mutate(across("ForecastUntil", as.Date)) %>%
  filter(!(City == "Houston" & ForecastFrom == "2023-09-29")) %>%
  distinct() %>%
  select(!"ForecastDate") %>%
  arrange(ForecastFrom)

forecasts_dated <- filter(forecasts_full, ForecastDate == "2023-09-28")
forecasts <- select(forecasts_dated, !"ForecastDate")

forecasts2 <- filter(forecasts_full, ForecastDate == "2023-09-29") %>%
  select(!"ForecastDate")
forecasts2a <- filter(forecasts_full, ForecastDate == "2023-09-29")
```

A slowly changing dimension is a concept in data warehousing which refers to data which may change over time, but at an irregular schedule.

## Type 1 and Type 2 history

For example, consider the following table of forecasts for a number of cities:
```{r forecasts, eval = requireNamespace("tidyverse")}
# Current date: 2023-09-28
forecasts
```

The following day, the forecasts will have changed, and — barring the occasional data hoarder — the existing data is no longer relevant.

In this example, most (if not all) of the values of the `Forecast` column will change with each regular update.
Putting it into other words, the table is a *snapshot*^[A *snapshot* is a static view of (part of) a database at a specific point in time] of forecasts at the last time of update.

The following day, the forecasts naturally change:
```{r forecasts2a, eval = requireNamespace("tidyverse")}
# Current date: 2023-09-29
forecasts2
```
We could choose to update the forecasts table so that it would always contain the current data.
This is what is referred to as Type 1 methodology [@Kimball2013].

Databases are thankfully a rather efficient way of storing and accessing data, so instead of discarding the values from the previous day, we append the new data to those of the previous day.
Also, in order to keep our data organized, we add a column with the date of the forecast, aptly named `ForecastDate`.

The full table of forecasts for the two days now looks like below, and we are slowly building a full history of forecasts:
```{r forecasts_full, eval = requireNamespace("tidyverse")}
forecasts_full
```
Managing historical data by inserting new data in this manner is often referred to as Type 2 methodology or Type 2 history.

Our table now provides much more information for the user through filtering:
```{r forecasts_full_examples, eval = requireNamespace("tidyverse")}

# Current forecasts
forecasts_full %>%
  slice_max(ForecastDate, n = 1) %>%
  select(!"ForecastDate")

# Forecasts for a given date
forecasts_full %>%
  filter(ForecastDate == "2023-09-28")

# Full history for a given city
forecasts_full %>%
  filter(City == "New York")
```

Now, we note that the forecast for Houston has not changed between the two days.

In order to keep our data as minimized as possible, we modify the table again, now expanding `ForecastDate` into `ForecastFrom` and `ForecastUntil`.

Our table of forecasts now looks like this:
```{r forecasts_scd, eval = requireNamespace("tidyverse")}
forecasts_scd
```
For now, the `ForecastUntil` value is set to `NA`, as it is not known when these rows will "expire" (if ever).
This also makes it easy to identify currently valid data.

Adding a new column to save a single row of data naturally seems a bit overkill, but as the number of rows in the data set increases indefinitely, this solutions scales much better.

## A "timeline of timelines"
Let's now introduce additional information and see how managing slowly changing dimensions enables us to easily navigate large amounts of data over large periods of time.

```{r adresses_setup, include = FALSE, eval = requireNamespace("tidyverse")}
addresses <- tibble::tibble(
  ID = c(1,
         2,
         1,
         1),
  GivenName = c("Alice",
                "Robert",
                "Alice",
                "Alice"),
  Surname = c("Doe",
              "Tables",
              "Doe",
              "Doe"),
  Address = c("Donut Plains 1",
              "Rainbow Road 8",
              "Donut Plains 1",
              "Rainbow Road 8"),
  MovedIn = c("1989-06-26",
              "1989-12-13",
              "1989-06-26",
              "2021-03-01"),
  MovedOut = c(NA,
               NA,
               "2021-03-01",
               NA),
  ValidFrom = c("1989-06-26",
                "1989-12-13",
                "2021-03-08",
                "2021-03-08"),
  ValidUntil = c("2021-03-08",
                 NA,
                 NA,
                 NA)
) %>%
  mutate(across(c("MovedIn", "MovedOut", "ValidFrom", "ValidUntil"), as.Date))


name_change_date <- as.Date("2023-08-28")
addresses2 <- addresses %>%
  filter(ID == 1) %>%
  {
    .data <- .

    .data %>%
      filter(Address == "Rainbow Road 8") %>%
      mutate(Surname = "Tables",
             ValidFrom = !!name_change_date) %>%
      bind_rows(.data)
  } %>%
  mutate(ValidUntil = as.Date(c(NA, !!name_change_date, NA, !!name_change_date), origin = "1970-01-01")) %>%
  bind_rows(filter(addresses,
                   ID == 2)) %>%
  arrange(ValidFrom) %>%
  distinct()
```

Imagine a town of several thousand citizens, with a town hall maintaining a civil registry of names and addresses of every citizen, updated daily with any changes submitted by the citizens, each of whom having an individual identification number.^[If this concept seems very familiar, you may have heard of [the Danish central civil registry](https://en.wikipedia.org/wiki/Personal_identification_number_(Denmark))]

The data is largely static, as a very small fraction of citizens move on any given day, but it is of interest to keep data relatively up-to-date.
This is where managing a slowly changing dimension becomes very powerful, compared to full incremental backups.

One day, Alice Doe meets [Robert "Bobby" Tables](https://xkcd.com/327/), and they move in together:

```{r addresses1, eval = requireNamespace("tidyverse")}
addresses
```

First thing to notice is that the registry is not updated in real-time, as citizens may have been late in registering a change of address.
This can be seen when comparing the values of `MovedIn` and `ValidFrom` for row 4.

When using Type 2 history, this feature is correctly replicated when reconstructing historical data:
```{r addresses1a, eval = requireNamespace("tidyverse")}
slice_timestamp <- "2021-03-02"

addresses %>%
  filter(ID == 1,
         ValidFrom < !!slice_timestamp,
         ValidUntil >= !!slice_timestamp | is.na(ValidUntil)) %>%
  select(!c("ValidFrom", "ValidUntil"))
```
In other words, even though Alice's address was subsequently updated in the registry, we can still see that she *was* registered as living in Donut Plains at this time.
This modeling of "timelines of timelines" is also called bitemporal modeling.

By now, things are going well between Alice and Robert; they get married, with Alice taking Robert's surname.
It is the same person that has lived with Robert, but as of the day of marriage, she has a different name:
```{r addresses2, eval = requireNamespace("tidyverse")}
filter(addresses2,
       ID == 1,
       Address == "Rainbow Road 8") %>%
  select(ID, GivenName, Surname, MovedIn, MovedOut, ValidFrom, ValidUntil)
```

This is now also reflected in the data; the `MovedIn` date is persistent across the date of the name change, only the `Surname` changes:
```{r addresses3, eval = requireNamespace("tidyverse")}
slice_timestamp <- "2022-03-04"

addresses2 %>%
  filter(Address == "Rainbow Road 8",
         is.na(MovedOut),
         ValidFrom < !!slice_timestamp,
         ValidUntil >= !!slice_timestamp | is.na(ValidUntil)) %>%
  select(ID, GivenName, Surname, MovedIn, MovedOut)

slice_timestamp <- "2023-09-29"

addresses2 %>%
  filter(Address == "Rainbow Road 8",
         is.na(MovedOut),
         ValidFrom < !!slice_timestamp,
         ValidUntil >= !!slice_timestamp | is.na(ValidUntil)) %>%
  select(ID, GivenName, Surname, MovedIn, MovedOut)
```

## Summary

By now, it is hopefully clear how managing a slowly changing dimension allows you to access data at any point in (tracked) time while potentially avoiding a lot of data redundancy.

You are now ready to get started with the `SCDB` package!

### References