vincentarelbundock

---
name: robot
description: Operate a local multi-agent development workflow built on Kata (issue/work ledger), kwt (git worktrees + tmux workspaces), Roborev (optional, operator-approved AI code review), and AgentsView (agent session history). Use when choosing or claiming work, creating a worktree for an issue, recording progress or evidence, reviewing commits, closing or abandoning work, running experiments, handing off between agents, supervising parallel worker agents, recovering after a crash, or when the user mentions kata, kwt, roborev, agentsview, /robot, "ready work", "claim", "worktree", "review loop", or "close with evidence".
---

# Robot: local agent development control plane

Four tools by Wes McKinney / kenn.io, plus Git. There is no other database. Every piece of state comes from the system that owns it:

| Question | Authority | Look at |

vincentarelbundock

#import "/.calepin/calepin.typ" as calepin

#set document(title: [Dynamic tabs from R])

= Dynamic tabs from R

The R function below accepts a named list and emits a
`calepin.elements.tabs` call at the location of the code chunk.

```r

vincentarelbundock

---
title: "Converting rtables to tinytable"
format: html
---

## Setup

```{r setup, message=FALSE, warning=FALSE}
library(tinytable)
library(rtables)

vincentarelbundock

Name: GNU General Public License
Abbrev: GPL
Version: 2
SSS: GPL-2
OSI: open (https://opensource.org/licenses/gpl-license)
FSF: free (https://www.gnu.org/licenses/license-list.html#GPLv2)
File: share/licenses/GPL-2
URL: https://www.r-project.org/Licenses/GPL-2
FOSS: yes
SPDX: GPL-2.0-only

vincentarelbundock

lambda1 <- 1.8
lambda2 <- 0.7
x <- 0:10
p1 <- dpois(x,lambda1)
p2 <- dpois(x,lambda2)
dp <- data.frame(x=c(x,x),p=c(p1,p2),lambda=rep(c(lambda1,lambda2),each=length(x)))
dp$lambda <- as.factor(dp$lambda)

plt(p ~ x,
    facet = ~lambda,

vincentarelbundock

library(marginaleffects)

# sanity checks: let's be super strict
sanity_vcov_unconditional <- function(p) {
  insight::check_if_installed("sandwich")
  cal <- attr(p, "call")

  callfun <- deparse(cal[[1]])
  if (!identical(callfun, "predictions")) {
    msg <- "`vcov='unconditional'` is only supported for `avg_predictions()` calls."

vincentarelbundock

<!DOCTYPE html> 
<html lang="en">

  <head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>tinytable_c80pijf8rq8f225cdrn6</title>
    <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.2/dist/css/bootstrap.min.css" rel="stylesheet">
    <script type="text/x-mathjax-config">

vincentarelbundock

# Downloads data here:

# https://academic.oup.com/ije/article/49/5/1719/5835351#226179197


#############################################################################
# TRIAL EMULATION: SURGERY WITHIN 6 MONTHS AMONG OLDER LUNG CANCER PATIENTS

# Author: Clemence Leyrat (from Camille Maringe's Stata code)
# Refactored by Vincent Arel-Bundock on 2023-12-08

vincentarelbundock

# Example of examining a continuous x categorical interaction using emmeans,
# and an attempt at doing the same using marginaleffects.
# Author: Cameron Patrick <cameron.patrick@unimelb.edu.au>

library(tidyverse)
library(emmeans)
library(marginaleffects)

# use the mtcars data, set up am as a factor
data(mtcars)

vincentarelbundock

``` r
library(marginaleffects)
library(modelsummary)

set.seed(20220722)
S <- diag(3)
S[1,2] <- S[2,1] <- 0.6
S[1,3] <- S[3,1] <- 0.6
data <- MASS::mvrnorm(1000, rep(0, 3), S) |> 
  as.data.frame() |>