index.qmd

---
title: "No-code data analysis and dashboards with the blockr ecosystem"
author: Nicolas Bennett
date: April 10, 2025
format:
  revealjs:
    progress: true
    controls: false
    transition: slide
---

# What is blockr about?

## Blocks! {transition="slide-in none-out"}

![ ](img/lego_blocks.jpeg){fig-align="center"}

## Blocks! {transition="none"}

![ ](img/lego_stack.jpeg){fig-align="center"}

## Blocks! {transition="none-in slide-out"}

![ ](img/lego_names.jpeg){fig-align="center"}

# Demo

## Analysis

How do different models of planes perform? What about Airbus vs. Boeing?

```{r}
#| include: false
library(dplyr)
library(ggplot2)
```

```{r}
#| echo: true
#| code-line-numbers: "1-4|6-9|11-15"
pln <- nycflights13::planes |>
  filter(seats >= 150) |>
  select(tailnum, manufacturer, model) |>
  mutate(manufacturer = sub(" INDUSTRIE$", "", manufacturer))

dat <- nycflights13::flights |>
  mutate(delay = -(arr_delay - dep_delay)) |>
  select(tailnum, delay) |>
  inner_join(pln, by = "tailnum")

dat <- dat |>
  summarize(count = n(), .by = c(manufacturer, model)) |>
  mutate(pos = rank(-count)) |>
  filter(pos <= 10) |>
  left_join(dat, by = c("manufacturer", "model"))
```

## Endpoints

::: columns
::: {.column width="50%"}

```{r}
#| echo: true
t.test(
  delay ~ manufacturer,
  data = dat
)
```

:::

::: {.column width="50%"}

```{r}
#| layout-ncol: 1
ggplot(dat, aes(delay, model)) +
  geom_boxplot()
ggplot(dat, aes(delay, color = manufacturer)) +
  geom_histogram()
```

:::
:::

## Interactive blockr dashboard

```{r}
#| eval: false
library(blockr.ui)
library(blockr.core)
library(blockr.dplyr)
library(blockr.ggplot)

run_demo_app(
  blocks = c(
    a1 = new_dataset_block("planes", "nycflights13", name = "Flights"),
    a2 = new_filter_block("seats >= 150"),
    a3 = new_select_block(c("tailnum", "manufacturer", "model")),
    a4 = new_mutate_block(
      list(manufacturer = "sub(\" INDUSTRIE$\", \"\", manufacturer)")
    ),
    b1 = new_dataset_block("flights", "nycflights13", name = "Planes"),
    b2 = new_mutate_block(list(delay = "-(arr_delay - dep_delay)")),
    b3 = new_select_block(c("tailnum", "delay")),
    c1 = new_join_block("inner", "tailnum"),
    c2 = new_summarize_block(
      list(count = "dplyr::n()"),
      c("manufacturer", "model")
    ),
    c3 = new_mutate_block(list(pos = "rank(-count)")),
    c4 = new_filter_block("pos <= 10"),
    c5 = new_join_block("left", c("manufacturer", "model")),
    d1 = new_boxplot_block("delay", "model"),
    d2 = new_histogram_block("delay", "manufacturer"),
    d3 = new_ttest_block("delay", "manufacturer")
  ),
  links = c(
    a1a2 = new_link("a1", "a2"),
    a2a3 = new_link("a2", "a3"),
    a3a4 = new_link("a3", "a4"),
    b1b2 = new_link("b1", "b2"),
    b2b3 = new_link("b2", "b3"),
    b3c1 = new_link("b3", "c1", "y"),
    a4c1 = new_link("a4", "c1", "x"),
    c1c2 = new_link("c1", "c2"),
    c2c3 = new_link("c2", "c3"),
    c3c4 = new_link("c3", "c4"),
    c4c5 = new_link("c4", "c5", "x"),
    c1c5 = new_link("c1", "c5", "y"),
    c5d1 = new_link("c5", "d1"),
    c5d2 = new_link("c5", "d2"),
    c5d3 = new_link("c5", "d3")
  ),
  stacks = list(
    a = paste0("a", 1:4),
    b = paste0("b", 1:3),
    c = paste0("c", 1:5),
    d = paste0("d", 1:3)
  )
)
```

![Graph view with blocks to the side](img/flights_1.png)

# Extensibility

## A framework?

- [blockr.core](https://github.com/cynkra/blockr.core) objects/tools for creating blocks/links/stacks and managing a board
- [blockr.ui](https://github.com/cynkra/blockr.ui) front end, extending the core functionality
- Block packages such as [blockr.dplyr](https://github.com/cynkra/blockr.dplyr)

We provide 2 APIs: one for adding blocks and one for customizing/extending the front end.

## Adding a new block (1/5) {transition="slide-in none-out"}

1. Specify a block constructor

```{r}
#| echo: true
#| code-line-numbers: "2|3-7|8-12|13"
new_lm_block <- function(response = character(), ...) {
  new_block(
    server = function(id, dat) {
      moduleServer(
        # ...
      )
    },
    ui = function(id) {
      tagList(
        # ...
      )
    },
    class = c("lm_block", "model_block"),
    ...
  )
}
```

## Adding a new block (1/5) {transition="none-in slide-out"}

1. Specify a block constructor

```{r}
#| echo: true
#| code-line-numbers: "29-31|8-9|11-15|18-23|24"
new_lm_block <- function(response = character(), ...) {
  new_block(
    server = function(id, dat) {
      moduleServer(
        id,
        function(input, output, session) {

          resp <- reactiveVal(response)
          observeEvent(input$resp, resp(input$resp))

          cols <- reactive(colnames(dat()))
          observeEvent(
            cols(),
            updateSelectInput(session, "resp", "Response", cols(), resp())
          )

          list(
            expr = reactive(
              bquote(
                stats::lm(.(y) ~ ., data = dat),
                list(y = as.name(resp()))
              )
            ),
            state = list(response = resp)
          )
        }
      )
    },
    ui = function(id) {
      selectInput(NS(id, "resp"), "Response", response, response)
    },
    class = c("lm_block", "model_block"),
    ...
  )
}
```

## Adding a new block (2/5)

2. (Optionally) provide some generic implementations for display

```{r}
#| echo: true
block_ui.model_block <- function(id, x, ...) {
  verbatimTextOutput(NS(id, "result"))
}

block_output.model_block <- function(x, result, session) {
  renderPrint(result)
}
```

## Adding a new block (3/5)

3. (Optionally) preview the block in a standalone app

```{r, eval = FALSE}
#| echo: true
serve(new_lm_block(), dat = datasets::attitude)
```

![Preview of an lm block](img/lm_block.png)

## Adding a new block (4/5)

4. Register the block for use in a dashboard

```{r, eval = FALSE}
#| echo: true
register_block(
  new_lm_block,
  name = "LM block",
  description = "Linear model block"
)
```

## Adding a new block (5/5)

Our current (WIP) block extension packages:

* [cynkra/blockr.dplyr](https://github.com/cynkra/blockr.dplyr)
* [cynkra/blockr.ai](https://github.com/cynkra/blockr.ai)
* [cynkra/blockr.io](https://github.com/cynkra/blockr.io)
* [cynkra/blockr.db](https://github.com/cynkra/blockr.db)
* [cynkra/blockr.gt](https://github.com/cynkra/blockr.gt)
* [cynkra/blockr.ggplot](https://github.com/cynkra/blockr.ggplot)

Your own?

## Extending the front-end (1/2)

A set of plugins (each a shiny module) can be user-supplied:

* `preserve_board`: Serialization/deserialization
* `manage_*`: Add/remove `blocks`, `links` and `stacks`
* `notify_user`: Handling of notifications
* `generate_code`: Code export
* `edit_*`: Per `block`/`stack` options

## Extending the front-end (2/2)

::: columns
::: {.column width="50%"}

```r
serve(
  new_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/core_ui.png)

:::

::: {.column width="50%"}

```r
serve(
  new_custom_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/custom_ui.png)

:::
:::

# Challenges

## Limitations of the shiny API

- Remove ([#2439](https://github.com/rstudio/shiny/issues/2439)) or reorder ([#4190](https://github.com/rstudio/shiny/issues/4190)) entries in a `reactiveValues` object: for variadic blocks (e.g. `rbind`) we need to be able to grow, shrink and reorder entries.
- Cleanup of dynamically created modules ([#2281](https://github.com/rstudio/shiny/issues/2281)): blocks (and stacks) are represented by modules which are dynamically added and removed.

In both cases we were able to come up with solutions (that circumvent the public API).

# Outro

## A collaborative effort

Joint work with David Granjon ([cynkra GmbH](https://cynkra.com)),

with contributions by John Coene ([The Y Company](https://the-y-company.com)), Karma Tarap ([Bristol-Myers Squibb](https://www.bms.com)) and Christoph Sax ([cynkra GmbH](https://cynkra.com)),

funded by [Bristol-Myers Squibb](https://www.bms.com).

## Further materials

- Prototype repo [BristolMyersSquibb/blockr](https://github.com/BristolMyersSquibb/blockr)
- Talk by David at useR 2024 in Salzburg ([slides](https://bristolmyerssquibb.github.io/useR2024))
- Workshop at "R in Pharma" ([Video](https://www.youtube.com/watch?v=PQvTQqcmadY))
- Discussion of prototype shortcomings and UI [wire-framing](https://excalidraw.com/#json=qYnQEvkzPgunMbErr4ulx,2c4c1kA16rY1Owhg54gPsQ)
- Re-write repo [cynkra/blockr.core](https://github.com/cynkra/blockr.core) with front-end package [cynkra/blockr.ui](https://github.com/cynkra/blockr.ui)

# Thanks for having me!

# Backup

## Interactive blockr dashboard {transition="slide-in none-out"}

![Dashboard builder with graph overview](img/flights_2.png)

## Interactive blockr dashboard {transition="none-in slide-out"}

![Dashboard view](img/flights_3.png)

## Extending the front-end (2/2) {transition="slide-in none-out"}

::: columns
::: {.column width="50%"}

```r
serve(
  new_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/core_block.png)

:::

::: {.column width="50%"}

```r
serve(
  new_custom_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/custom_block.png)

:::
:::

## Extending the front-end (2/2) {transition="none"}

::: columns
::: {.column width="50%"}

```r
serve(
  new_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/core_link.png)

:::

::: {.column width="50%"}

```r
serve(
  new_custom_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/custom_link.png)

:::
:::

## Extending the front-end (2/2) {transition="none-in slide-out"}

::: columns
::: {.column width="50%"}

```r
serve(
  new_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/core_stack.png)

:::

::: {.column width="50%"}

```r
serve(
  new_custom_board(
    c(a = new_dataset_block("mtcars"))
  )
)
```

![](img/custom_stack.png)

:::
:::