From ff5be1e06a682719f739a1624e30204b169ba04e Mon Sep 17 00:00:00 2001
From: Andrew Johnson <andrew.johnson@arjohnsonau.com>
Date: Mon, 23 Oct 2023 09:26:28 +0300
Subject: [PATCH] Update aaltobda install instructions

---
 FAQ.Rmd                               |   6 +-
 assignments/assignment6.qmd           |  83 +++++++++--------
 assignments/includes/_general_info.md |   3 +-
 assignments/simple_template3.qmd      |  92 +++++++++----------
 assignments/template2.qmd             |  48 +++++-----
 assignments/template3.qmd             | 126 +++++++++++++-------------
 assignments/template4.qmd             |  38 ++++----
 assignments/template5.qmd             |  44 +++++----
 assignments/template6.qmd             |  58 ++++++------
 9 files changed, 241 insertions(+), 257 deletions(-)

diff --git a/FAQ.Rmd b/FAQ.Rmd
index cb0c8250..ea0ca03c 100644
--- a/FAQ.Rmd
+++ b/FAQ.Rmd
@@ -120,11 +120,9 @@ install.packages(c("MASS", "bayesplot", "brms", "cmdstanr ", "dplyr", "gganimate
 
 The course has its own R package `aaltobda` with data and
 functionality to simplify coding. `aaltobda` has been pre-installed in
-Aalto JupyterHub. To install the package to your own computer just run the following
-(upgrade=”never” skips question about updating other packages):
+Aalto JupyterHub. To install the package to your own computer just run the following:
 
-1. `install.packages("remotes")`
-2. `remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never", dependencies=TRUE)`
+1. `install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))`
 
 If during the course there is announcement that `aaltobda` has been
 updated (e.g. some error has been fixed), you can get the latest
diff --git a/assignments/assignment6.qmd b/assignments/assignment6.qmd
index e44077eb..3082814c 100644
--- a/assignments/assignment6.qmd
+++ b/assignments/assignment6.qmd
@@ -1,33 +1,33 @@
 ---
 title: "Assignment 6"
 author: "Aki Vehtari et al."
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
 editor: source
 filters:
-  - includes/assignments.lua 
-  - includes/include-code-files.lua 
+  - includes/assignments.lua
+  - includes/include-code-files.lua
 ---
 
 # General information
 
 **The maximum amount of points from this assignment is 6.**
 
-We have prepared a **quarto template specific to this assignment ([html](template6.html), [qmd](https://avehtari.github.io/BDA_course_Aalto/assignments/template6.qmd), [pdf](template6.pdf))** to help you get started. 
+We have prepared a **quarto template specific to this assignment ([html](template6.html), [qmd](https://avehtari.github.io/BDA_course_Aalto/assignments/template6.qmd), [pdf](template6.pdf))** to help you get started.
 
 
 ::: {.aalto}
 We recommend you use [jupyter.cs.aalto.fi](https://jupyter.cs.aalto.fi) or the [docker container](docker.html).
 :::
 
-::: {.hint}             
-**Reading instructions:** 
+::: {.hint}
+**Reading instructions:**
 
 - [**The reading instructions for BDA3 Chapter 10**](../BDA3_notes.html#ch10).
 - [**The reading instructions for BDA3 Chapter 11**](../BDA3_notes.html#ch11).
@@ -44,16 +44,15 @@ We recommend you use [jupyter.cs.aalto.fi](https://jupyter.cs.aalto.fi) or the [
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-tip collapse=true}
- 
-## Setup 
+
+## Setup
 
 JupyterHub has all the needed packages pre-installed.
 
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -129,7 +128,7 @@ p(y|x,\alpha,\beta,\sigma) &= p_\mathrm{normal}(y|\alpha + \beta x, \sigma)
       & \text{(normal likelihood)} &\text{.}
 \end{aligned}
 $$
-In both the statistical model above and in the Stan model below, $x \in \mathbb{R}^N$ and $y \in \mathbb{R}^N$ are vectors of the covariates / predictors (the assignment number) and vectors of the observation (proportions of students who have handed in the respective assignment). $\alpha \in \mathbb{R}$ is the unknown scalar intercept, $\beta \in \mathbb{R}$ is the unknown scalar slope and $\sigma \in \mathbb{R}_{>0}$ is the unknown scalar observation standard deviation. The statistical model further implies 
+In both the statistical model above and in the Stan model below, $x \in \mathbb{R}^N$ and $y \in \mathbb{R}^N$ are vectors of the covariates / predictors (the assignment number) and vectors of the observation (proportions of students who have handed in the respective assignment). $\alpha \in \mathbb{R}$ is the unknown scalar intercept, $\beta \in \mathbb{R}$ is the unknown scalar slope and $\sigma \in \mathbb{R}_{>0}$ is the unknown scalar observation standard deviation. The statistical model further implies
 $$
 p(y_\mathrm{pred.}|x_\mathrm{pred.},\alpha,\beta,\sigma) = p_\mathrm{normal}(y_\mathrm{pred.}|\alpha + \beta x_\mathrm{pred.}, \sigma)
 $$
@@ -140,23 +139,23 @@ You can download [the broken stan file from github](./additional_files/assignmen
 ```{.stan}
 data { #<1>
     // number of data points
-    int<lower=0> N; 
+    int<lower=0> N;
     // covariate / predictor
-    vector[N] x; 
+    vector[N] x;
     // observations
-    vector[N] y; 
+    vector[N] y;
     // number of covariate values to make predictions at
     int<lower=0> no_predictions;
     // covariate values to make predictions at
-    vector[no_predictions] x_predictions; 
+    vector[no_predictions] x_predictions;
 } #<1>
 parameters { #<2>
     // intercept
-    real alpha; 
+    real alpha;
     // slope
-    real beta; 
+    real beta;
     // the standard deviation should be constrained to be positive
-    real<upper=0> sigma; 
+    real<upper=0> sigma;
 } #<2>
 transformed parameters { #<3>
     // deterministic transformation of parameters and data
@@ -164,7 +163,7 @@ transformed parameters { #<3>
 } #<3>
 model { #<4>
     // observation model / likelihood
-    y ~ normal(mu, sigma); 
+    y ~ normal(mu, sigma);
 } #<4>
 generated quantities { #<5>
     // compute the means for the covariate values at which to make predictions
@@ -200,7 +199,7 @@ Find the ***three mistakes*** in the code and fix them. Report the original mist
 
 ::: {.hint}
 You may find some of the mistakes in the code using Stan syntax checker. If you copy the Stan code to a file ending `.stan` and open it in RStudio (you can also choose from RStudio menu File$\rightarrow$New File$\rightarrow$Stan file to create a new Stan file), the editor will show you some syntax errors. More syntax errors might be detected by clicking `Check' in the bar just above the Stan file in the RStudio editor. Note that some of the errors in the presented Stan code may not be syntax errors.
-::: 
+:::
 
 
 
@@ -218,7 +217,7 @@ The author runs the corrected Stan file using the following R code and plots the
 #| warning: false
 # These are our observations y: the proportion of students handing in each assignment (1-8),
 # sorted by year (row-wise) and assignment (column-wise).
-# While the code suggest a matrix structure, 
+# While the code suggest a matrix structure,
 # the result will actually be a vector of length N = no_years * no_assignments
 propstudents<-c(c(176, 174, 158, 135, 138, 129, 126, 123)/176,
                 c(242, 212, 184, 177, 174, 172, 163, 156)/242,
@@ -228,7 +227,7 @@ propstudents<-c(c(176, 174, 158, 135, 138, 129, 126, 123)/176,
 # These are our predictors x: for each observation, the corresponding assignment number.
 assignment <- rep(1:8, 5)
 # These are in some sense our test data: the proportion of students handing in the last assignment (9),
-# sorted by year. 
+# sorted by year.
 # Usually, we would not want to split our data like that and instead
 # use e.g. Leave-One-Out Cross-Validation (LOO-CV, see e.g. http://mc-stan.org/loo/index.html)
 # to evaluate model performance.
@@ -246,8 +245,8 @@ model_data = list(N=length(assignment),
 ```
 **Sampling from the posterior distribution happens here**:
 ```{r}
-#| warning: false 
-# This reads the file at the specified path and tries to compile it. 
+#| warning: false
+# This reads the file at the specified path and tries to compile it.
 # If it fails, an error is thrown.
 retention_model = cmdstan_model("./additional_files/assignment6_linear_model.stan")
 # This "out <- capture.output(...)" construction suppresses output from cmdstanr
@@ -263,21 +262,21 @@ out <- capture.output(
 # This extracts the draws from the sampling result as a data.frame.
 draws_df = fit$draws(format="draws_df")
 
-# This does some data/draws wrangling to compute the 5, 50 and 95 percentiles of 
-# the mean at the specified covariate values (x_predictions). 
+# This does some data/draws wrangling to compute the 5, 50 and 95 percentiles of
+# the mean at the specified covariate values (x_predictions).
 # It can be instructive to play around with each of the data processing steps
 # to find out what each step does, e.g. by removing parts from the back like "|>  gather(pct,y,-x)"
 # and printing the resulting data.frame.
-mu_quantiles_df = draws_df |> 
-      subset_draws(variable = c("mu_pred")) |> 
-      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |> 
-      mutate(x = 1:9) |> 
+mu_quantiles_df = draws_df |>
+      subset_draws(variable = c("mu_pred")) |>
+      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |>
+      mutate(x = 1:9) |>
       pivot_longer(c(q5, q50, q95), names_to = c("pct"))
 # Same as above, but for the predictions.
-y_quantiles_df = draws_df |> 
-      subset_draws(variable = c("y_pred")) |> 
-      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |> 
-      mutate(x = 1:9) |> 
+y_quantiles_df = draws_df |>
+      subset_draws(variable = c("y_pred")) |>
+      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |>
+      mutate(x = 1:9) |>
       pivot_longer(c(q5, q50, q95), names_to = c("pct"))
 ```
 
@@ -290,14 +289,14 @@ y_quantiles_df = draws_df |>
 #| label: fig-posterior
 #| fig-cap: Describe me in your submission!
 ggplot() +
-  # scatter plot of the training data:  
+  # scatter plot of the training data:
   geom_point(
-    aes(x, y, color=assignment), 
+    aes(x, y, color=assignment),
     data=data.frame(x=assignment, y=propstudents, assignment="1-8")
 ) +
   # scatter plot of the test data:
   geom_point(
-    aes(x, y, color=assignment), 
+    aes(x, y, color=assignment),
     data=data.frame(x=no_assignments, y=propstudents9, assignment="9")
 ) +
   # you have to tell us what this plots:
@@ -340,7 +339,7 @@ Based on the above plot, answer the following questions:
 - What is the general trend of student retention as measured by assignment submissions?
 - Given a model fitted to the submission data for assignments 1-8, does it do a good job predicting the proportion of students who submit the final 9th assignment?
 - Name one different modeling choice you could make to improve the prediction.
-::: 
+:::
 
 
 ::: {.rubric}
@@ -379,7 +378,7 @@ Based on the above plot, answer the following questions:
 * Has at least one way to improve the model been mentioned (E.g. **...** or **...**)?
 :::
 
-# Generalized linear model: Bioassay with Stan (4 points) 
+# Generalized linear model: Bioassay with Stan (4 points)
 
 Replicate the computations for the bioassay example of section 3.7
 (BDA3) using Stan.
@@ -460,13 +459,13 @@ any problems in setting it up or using it. Please report,
 :::
 
 
- 
+
 
 ::: {.rubric}
 * Is the Stan model code included?
     * No
     * Yes
-* Does the implemented Stan-model seem to be working? 
+* Does the implemented Stan-model seem to be working?
     * No implementation
     * Model implemented but results not visualized/reported
     * Model implemented, but the results seem weird
diff --git a/assignments/includes/_general_info.md b/assignments/includes/_general_info.md
index 547a5ad0..9b4ec664 100644
--- a/assignments/includes/_general_info.md
+++ b/assignments/includes/_general_info.md
@@ -20,8 +20,7 @@
   To install the package on your own system, run
   the following code (upgrade=\"never\" skips question about updating other packages):
 ```{.r}
-install.packages("remotes")
-remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
 ```
 -   Many of the exercises can be checked automatically using the R
     package `markmyassignment` (pre-installed in JupyterHub).
diff --git a/assignments/simple_template3.qmd b/assignments/simple_template3.qmd
index 314ccb14..0900b757 100644
--- a/assignments/simple_template3.qmd
+++ b/assignments/simple_template3.qmd
@@ -1,15 +1,15 @@
 ---
 title: "Assignment 3"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
@@ -22,37 +22,36 @@ editor: source
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
 This is the template for [assignment 3](assignment3.html). You can download the qmd-files ([full](https://avehtari.github.io/BDA_course_Aalto/assignments/template3.qmd), [simple](https://avehtari.github.io/BDA_course_Aalto/assignments/simple_template3.qmd)) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 The following will set-up [`markmyassignment`](https://github.com/MansMeg/markmyassignment) to check your functions at the end of the notebook:
- 
-```{r} 
+
+```{r}
 if(!require(markmyassignment)){
     install.packages("markmyassignment")
     library(markmyassignment)
 }
 assignment_path = paste("https://github.com/avehtari/BDA_course_Aalto/",
 "blob/master/assignments/tests/assignment3.yml", sep="")
-set_assignment(assignment_path)    
-```       
+set_assignment(assignment_path)
+```
 
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -70,7 +69,7 @@ if(!require(latex2exp)){
 
 
 
-# Inference for normal mean and deviation (3 points) 
+# Inference for normal mean and deviation (3 points)
 
 Loading the library and the data.
 ``` {r}
@@ -90,17 +89,17 @@ windshieldy_test <- c(13.357, 14.928, 14.896, 14.820)
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
 ## (b)
 
 
-Write your answers and code here! 
+Write your answers and code here!
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
-```{r} 
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum()
 # and qtnew(), dtnew() (from aaltobda)
 
@@ -108,13 +107,13 @@ mu_point_est <- function(data) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     14.5
-    
+
 }
 mu_interval <- function(data, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(13.3, 15.7)
-    
+
 }
 ```
 
@@ -125,13 +124,13 @@ You can plot the density as below if you implement `mu_pdf` to compute the PDF o
 #| fig-cap: PDF of the posterior $p(\mu|y)$ of the average hardness $\mu$
 mu_pdf <- function(data, x){
     # Compute necessary parameters here.
-    # These are the correct parameters for `windshieldy_test` 
+    # These are the correct parameters for `windshieldy_test`
     # with the provided uninformative prior.
     df = 3
     location = 14.5
     scale = 0.3817557
     # Use the computed parameters as below to compute the PDF:
-     
+
     dtnew(x, df, location, scale)
 }
 
@@ -140,8 +139,8 @@ lower_x = x_interval[1]
 upper_x = x_interval[2]
 x = seq(lower_x, upper_x, length.out=1000)
 plot(
-    x, mu_pdf(windshieldy1, x), type="l", 
-    xlab=TeX(r'(average hardness $\mu$)'), 
+    x, mu_pdf(windshieldy1, x), type="l",
+    xlab=TeX(r'(average hardness $\mu$)'),
     ylab=TeX(r'(PDF of the posterior $p(\mu|y)$)')
 )
 ```
@@ -154,7 +153,7 @@ plot(
 Write your answers and code here!
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
-```{r} 
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum()
 # and qtnew(), dtnew() (from aaltobda)
 
@@ -162,13 +161,13 @@ mu_pred_point_est <- function(data) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     14.5
-    
+
 }
 mu_pred_interval <- function(data, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(11.8, 17.2)
-    
+
 }
 ```
 
@@ -179,13 +178,13 @@ You can plot the density as below if you implement `mu_pred_pdf` to compute the
 #| fig-cap: PDF of the posterior predictive $p(\tilde{y}|y)$ of a new hardness observation $\tilde{y}$
 mu_pred_pdf <- function(data, x){
     # Compute necessary parameters here.
-    # These are the correct parameters for `windshieldy_test` 
+    # These are the correct parameters for `windshieldy_test`
     # with the provided uninformative prior.
     df = 3
     location = 14.5
     scale = 0.8536316
     # Use the computed parameters as below to compute the PDF:
-     
+
     dtnew(x, df, location, scale)
 }
 
@@ -194,19 +193,19 @@ lower_x = x_interval[1]
 upper_x = x_interval[2]
 x = seq(lower_x, upper_x, length.out=1000)
 plot(
-    x, mu_pred_pdf(windshieldy1, x), type="l", 
-    xlab=TeX(r'(new hardness observation $\tilde{y}$)'), 
+    x, mu_pred_pdf(windshieldy1, x), type="l",
+    xlab=TeX(r'(new hardness observation $\tilde{y}$)'),
     ylab=TeX(r'(PDF of the posterior predictive $p(\tilde{y}|y)$)')
 )
 ```
 
 
-# Inference for the difference between proportions (3 points) 
+# Inference for the difference between proportions (3 points)
 
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
@@ -226,20 +225,20 @@ p1 = rbeta(ndraws, 10, 90)
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
 
-```{r} 
+```{r}
 # Useful function: mean(), quantile()
 
 posterior_odds_ratio_point_est <- function(p0, p1) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     2.650172
-    
+
 }
 posterior_odds_ratio_interval <- function(p0, p1, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(0.6796942,7.3015964)
-    
+
 }
 ```
 
@@ -250,10 +249,10 @@ posterior_odds_ratio_interval <- function(p0, p1, prob = 0.95) {
 ## (c)
 
 
-Write your answers and code here! 
+Write your answers and code here!
 
 
-# Inference for the difference between normal means (3 points) 
+# Inference for the difference between normal means (3 points)
 
 Loading the library and the data.
 ``` {r}
@@ -268,15 +267,15 @@ head(windshieldy2)
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
 ## (b)
 
 
-Write your answers and code here! 
-```{r} 
+Write your answers and code here!
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum(),
 # rtnew() (from aaltobda), quantile() and hist().
 ```
@@ -286,7 +285,7 @@ Write your answers and code here!
 ## (c)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
@@ -298,15 +297,14 @@ Write your answers here!
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 The following will check the functions for which `markmyassignment` has been set up:
- 
-```{r}  
-mark_my_assignment()    
-```      
+
+```{r}
+mark_my_assignment()
+```
 
 :::
 ::::
-
diff --git a/assignments/template2.qmd b/assignments/template2.qmd
index 87c3a6e2..7f1d1002 100644
--- a/assignments/template2.qmd
+++ b/assignments/template2.qmd
@@ -1,23 +1,23 @@
 ---
 title: "Assignment 2"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
     code-annotations: none
-    include-in-header: 
+    include-in-header:
       text: |
-        % You can add TeX macros here for PDF, 
+        % You can add TeX macros here for PDF,
         % see https://quarto.org/docs/output-formats/pdf-basics.html#latex-includes
-        \newcommand{\BetaDist}{\mathrm{Beta}} 
+        \newcommand{\BetaDist}{\mathrm{Beta}}
 editor: source
 ---
 
@@ -33,35 +33,34 @@ $$
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
 This is the template for [assignment 2](assignment2.html). You can download the [qmd-file](https://avehtari.github.io/BDA_course_Aalto/assignments/template2.qmd) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 The following will set-up `markmyassignment` to check your functions at the end of the notebook:
- 
-```{r} 
-library(markmyassignment) 
+
+```{r}
+library(markmyassignment)
 assignment_path = paste("https://github.com/avehtari/BDA_course_Aalto/",
 "blob/master/assignments/tests/assignment2.yml", sep="")
-set_assignment(assignment_path)    
-```       
+set_assignment(assignment_path)
+```
 
 The following installs the `aaltobda` package:
 ```{r}
 #| cache: true
 # Caching should be fine here
-install.packages("remotes")
-remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
 ```
 
 :::
@@ -93,7 +92,7 @@ algae_test <- c(0, 1, 1, 0, 0, 0)
 
 Write the likelihood, the prior and the posterior here!
 ```{r}
-# These are not the actual values for the posterior! 
+# These are not the actual values for the posterior!
 # You will have to compute those from the data!
 posterior_alpha = 2
 posterior_beta = 10
@@ -112,7 +111,7 @@ This template defines a `\BetaDist` TeX command which renders `$\BetaDist(1,2)$`
 
 
 Keep the below name and format for the functions to work with `markmyassignment`:
-```{r} 
+```{r}
 # Useful function: qbeta()
 
 beta_point_est <- function(prior_alpha, prior_beta, data) {
@@ -173,15 +172,14 @@ Plot the PDFs here. Explain shortly what you do.
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 The following will check the functions for which `markmyassignment` has been set up:
- 
-```{r}  
-mark_my_assignment()    
-```      
+
+```{r}
+mark_my_assignment()
+```
 
 :::
 ::::
-
diff --git a/assignments/template3.qmd b/assignments/template3.qmd
index cb57c975..12249f72 100644
--- a/assignments/template3.qmd
+++ b/assignments/template3.qmd
@@ -1,15 +1,15 @@
 ---
 title: "Assignment 3"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
@@ -22,37 +22,36 @@ editor: source
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
 This is the template for [assignment 3](assignment3.html). You can download the qmd-files ([full](https://avehtari.github.io/BDA_course_Aalto/assignments/template3.qmd), [simple](https://avehtari.github.io/BDA_course_Aalto/assignments/simple_template3.qmd)) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 The following will set-up [`markmyassignment`](https://github.com/MansMeg/markmyassignment) to check your functions at the end of the notebook:
- 
-```{r} 
+
+```{r}
 if(!require(markmyassignment)){
     install.packages("markmyassignment")
     library(markmyassignment)
 }
 assignment_path = paste("https://github.com/avehtari/BDA_course_Aalto/",
 "blob/master/assignments/tests/assignment3.yml", sep="")
-set_assignment(assignment_path)    
-```       
+set_assignment(assignment_path)
+```
 
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -77,7 +76,7 @@ if(!require(latex2exp)){
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 *This block showcases advanced tools, which you will be allowed and expected to use after this assignment.*
@@ -119,9 +118,9 @@ if(!require(ggdist)){
 
 :::
 ::::
-    
 
-# Inference for normal mean and deviation (3 points) 
+
+# Inference for normal mean and deviation (3 points)
 
 Loading the library and the data.
 ``` {r}
@@ -141,17 +140,17 @@ windshieldy_test <- c(13.357, 14.928, 14.896, 14.820)
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
 ## (b)
 
 
-Write your answers and code here! 
+Write your answers and code here!
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
-```{r} 
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum()
 # and qtnew(), dtnew() (from aaltobda)
 
@@ -159,13 +158,13 @@ mu_point_est <- function(data) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     14.5
-    
+
 }
 mu_interval <- function(data, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(13.3, 15.7)
-    
+
 }
 ```
 
@@ -176,13 +175,13 @@ You can plot the density as below if you implement `mu_pdf` to compute the PDF o
 #| fig-cap: PDF of the posterior $p(\mu|y)$ of the average hardness $\mu$
 mu_pdf <- function(data, x){
     # Compute necessary parameters here.
-    # These are the correct parameters for `windshieldy_test` 
+    # These are the correct parameters for `windshieldy_test`
     # with the provided uninformative prior.
     df = 3
     location = 14.5
     scale = 0.3817557
     # Use the computed parameters as below to compute the PDF:
-     
+
     dtnew(x, df, location, scale)
 }
 
@@ -191,8 +190,8 @@ lower_x = x_interval[1]
 upper_x = x_interval[2]
 x = seq(lower_x, upper_x, length.out=1000)
 plot(
-    x, mu_pdf(windshieldy1, x), type="l", 
-    xlab=TeX(r'(average hardness $\mu$)'), 
+    x, mu_pdf(windshieldy1, x), type="l",
+    xlab=TeX(r'(average hardness $\mu$)'),
     ylab=TeX(r'(PDF of the posterior $p(\mu|y)$)')
 )
 ```
@@ -205,7 +204,7 @@ plot(
 Write your answers and code here!
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
-```{r} 
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum()
 # and qtnew(), dtnew() (from aaltobda)
 
@@ -213,13 +212,13 @@ mu_pred_point_est <- function(data) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     14.5
-    
+
 }
 mu_pred_interval <- function(data, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(11.8, 17.2)
-    
+
 }
 ```
 
@@ -230,13 +229,13 @@ You can plot the density as below if you implement `mu_pred_pdf` to compute the
 #| fig-cap: PDF of the posterior predictive $p(\tilde{y}|y)$ of a new hardness observation $\tilde{y}$
 mu_pred_pdf <- function(data, x){
     # Compute necessary parameters here.
-    # These are the correct parameters for `windshieldy_test` 
+    # These are the correct parameters for `windshieldy_test`
     # with the provided uninformative prior.
     df = 3
     location = 14.5
     scale = 0.8536316
     # Use the computed parameters as below to compute the PDF:
-     
+
     dtnew(x, df, location, scale)
 }
 
@@ -245,19 +244,19 @@ lower_x = x_interval[1]
 upper_x = x_interval[2]
 x = seq(lower_x, upper_x, length.out=1000)
 plot(
-    x, mu_pred_pdf(windshieldy1, x), type="l", 
-    xlab=TeX(r'(new hardness observation $\tilde{y}$)'), 
+    x, mu_pred_pdf(windshieldy1, x), type="l",
+    xlab=TeX(r'(new hardness observation $\tilde{y}$)'),
     ylab=TeX(r'(PDF of the posterior predictive $p(\tilde{y}|y)$)')
 )
 ```
 
 
-# Inference for the difference between proportions (3 points) 
+# Inference for the difference between proportions (3 points)
 
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
@@ -277,20 +276,20 @@ p1 = rbeta(ndraws, 10, 90)
 
 **Keep the below name and format for the functions to work with `markmyassignment`:**
 
-```{r} 
+```{r}
 # Useful function: mean(), quantile()
 
 posterior_odds_ratio_point_est <- function(p0, p1) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     2.650172
-    
+
 }
 posterior_odds_ratio_interval <- function(p0, p1, prob = 0.95) {
     # Do computation here, and return as below.
     # This is the correct return value for the test data provided above.
     c(0.6796942,7.3015964)
-    
+
 }
 ```
 
@@ -298,13 +297,13 @@ posterior_odds_ratio_interval <- function(p0, p1, prob = 0.95) {
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-tip collapse=false}
- 
+
 ## Showcase: advanced tools (`posterior`'s `rvar`, `ggdist`'s `stat_dotsinterval`)
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 *This block showcases advanced tools, which you will be allowed and expected to use after this assignment.*
@@ -313,15 +312,15 @@ posterior_odds_ratio_interval <- function(p0, p1, prob = 0.95) {
 
 
 
-The `posterior` package's random variable datatype `rvar` is a 
+The `posterior` package's random variable datatype `rvar` is a
 ["sample-based representation of random variables"](https://mc-stan.org/posterior/articles/rvar.html#:~:text=sample%2Dbased%20representation%20of%20random%20variables)
-which makes handling of random samples (of draws) such as the ones contained in the above variables `p0` and `p1` easier. 
-[By default, it prints as the mean and standard deviation of the draws](https://mc-stan.org/posterior/articles/rvar.html#:~:text=The%20default%20display%20of%20an%20rvar%20shows%20the%20mean%20and%20standard%20deviation%20of%20each%20element%20of%20the%20array.), **such that `rvar(p0)` prints as `r rvar(p0)` and `rvar(p1)` prints as `r rvar(p1)`**. 
+which makes handling of random samples (of draws) such as the ones contained in the above variables `p0` and `p1` easier.
+[By default, it prints as the mean and standard deviation of the draws](https://mc-stan.org/posterior/articles/rvar.html#:~:text=The%20default%20display%20of%20an%20rvar%20shows%20the%20mean%20and%20standard%20deviation%20of%20each%20element%20of%20the%20array.), **such that `rvar(p0)` prints as `r rvar(p0)` and `rvar(p1)` prints as `r rvar(p1)`**.
 
 The datatype is ["designed to [...] be able to be used inside `data.frame()`s and `tibble()`s, and to be used with distribution visualizations in the ggdist package."](https://mc-stan.org/posterior/articles/rvar.html#:~:text=designed%20to%20interoperate%20with%20vectorized%20distributions%20in%20the%20distributional%20package%2C%20to%20be%20able%20to%20be%20used%20inside%20data.frame()s%20and%20tibble()s%2C%20and%20to%20be%20used%20with%20distribution%20visualizations%20in%20the%20ggdist%20package.)
-The code below sets up an [R `data.frame()`](https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/data.frame) with the draws in `p0` and `p1` wrapped in an `rvar`, and uses that data frame to visualize the draws using [`ggdist`](https://mjskay.github.io/ggdist/index.html), an R package building on [`ggplot2`](https://ggplot2.tidyverse.org/) and ["designed for both frequentist and Bayesian uncertainty visualization"](https://mjskay.github.io/ggdist/index.html#:~:text=designed%20for%20both%20frequentist%20and%20Bayesian%20uncertainty%20visualization). 
+The code below sets up an [R `data.frame()`](https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/data.frame) with the draws in `p0` and `p1` wrapped in an `rvar`, and uses that data frame to visualize the draws using [`ggdist`](https://mjskay.github.io/ggdist/index.html), an R package building on [`ggplot2`](https://ggplot2.tidyverse.org/) and ["designed for both frequentist and Bayesian uncertainty visualization"](https://mjskay.github.io/ggdist/index.html#:~:text=designed%20for%20both%20frequentist%20and%20Bayesian%20uncertainty%20visualization).
 
-The below plot, @fig-showcase-probabilities uses `ggdist`'s [`stat_dotsinterval()`](https://mjskay.github.io/ggdist/articles/dotsinterval.html), which by default visualizes  
+The below plot, @fig-showcase-probabilities uses `ggdist`'s [`stat_dotsinterval()`](https://mjskay.github.io/ggdist/articles/dotsinterval.html), which by default visualizes
 
 * [an `rvar`'s median and central 66% and 95% intervals](https://mjskay.github.io/ggdist/reference/stat_dotsinterval.html#:~:text=point_interval%20%3D%20%22median_qi%22%2C%0A%20%20.width%20%3D%20c(0.66%2C%200.95)%2C) using a black dot and lines of varying thicknesses as when using `ggdist`'s [`stat_pointinterval()`](https://mjskay.github.io/ggdist/reference/stat_pointinterval.html#examples) and
 * an `rvar`'s draws using grey dots as when using `ggdist`'s [`stat_dots()`](https://mjskay.github.io/ggdist/reference/stat_dots.html#examples):
@@ -334,19 +333,19 @@ r1 = rvar(p1)
 ggplot(data.frame(
     rv_name=c("control", "treatment"), rv=c(r0, r1)
 )) +
-    aes(xdist=rv, y=rv_name) + 
-    labs(x="probabilities of death", y="patient group") + 
+    aes(xdist=rv, y=rv_name) +
+    labs(x="probabilities of death", y="patient group") +
     stat_dotsinterval()
 ```
 
-`rvar`s make it easy to compute functions of random variables, such as 
+`rvar`s make it easy to compute functions of random variables, such as
 
 * differences, e.g. $p_0 - p_1$: `r0 - r1` computes an `rvar` which prints as `r r0 - r1`, indicating the **sample mean** and the **sample standard deviation** of the difference of the probabilities of death,
 * products, e.g. $p_0 \, p_1$: `r0 * r1` computes an `rvar` which prints as `r r0 * r1` which in this case
-has no great interpretation, or 
+has no great interpretation, or
 * the odds ratios needed in task 3.b).
 
-Below, in @fig-showcase-odds-ratios, we compute the odds ratios using the `rvar`s and visualize its median, central intervals and draws, as above in @fig-showcase-probabilities: 
+Below, in @fig-showcase-odds-ratios, we compute the odds ratios using the `rvar`s and visualize its median, central intervals and draws, as above in @fig-showcase-probabilities:
 ```{r}
 #| label: fig-showcase-odds-ratios
 #| fig-cap: Odds ratios of the two patient groups.
@@ -354,8 +353,8 @@ rodds_ratio = (r1/(1-r1))/(r0/(1-r0))
 ggplot(data.frame(
     rv=c(rodds_ratio)
 )) +
-    aes(xdist=rv) + 
-    labs(x="odds ratio", y="relative amount of draws") + 
+    aes(xdist=rv) +
+    labs(x="odds ratio", y="relative amount of draws") +
     stat_dotsinterval()
 ```
 
@@ -370,16 +369,16 @@ You can use @fig-showcase-odds-ratios to visually check whether the answers you
 
 :::
 ::::
-    
+
 
 
 ## (c)
 
 
-Write your answers and code here! 
+Write your answers and code here!
 
 
-# Inference for the difference between normal means (3 points) 
+# Inference for the difference between normal means (3 points)
 
 Loading the library and the data.
 ``` {r}
@@ -394,15 +393,15 @@ head(windshieldy2)
 ## (a)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
 ## (b)
 
 
-Write your answers and code here! 
-```{r} 
+Write your answers and code here!
+```{r}
 # Useful functions: mean(), length(), sqrt(), sum(),
 # rtnew() (from aaltobda), quantile() and hist().
 ```
@@ -412,7 +411,7 @@ Write your answers and code here!
 ## (c)
 
 
-Write your answers here! 
+Write your answers here!
 
 
 
@@ -424,15 +423,14 @@ Write your answers here!
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 The following will check the functions for which `markmyassignment` has been set up:
- 
-```{r}  
-mark_my_assignment()    
-```      
+
+```{r}
+mark_my_assignment()
+```
 
 :::
 ::::
-
diff --git a/assignments/template4.qmd b/assignments/template4.qmd
index de83f7bf..952c159d 100644
--- a/assignments/template4.qmd
+++ b/assignments/template4.qmd
@@ -1,15 +1,15 @@
 ---
 title: "Assignment 4"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
@@ -22,37 +22,36 @@ editor: source
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
 This is the template for [assignment 4](assignment4.html). You can download the [qmd-file](https://avehtari.github.io/BDA_course_Aalto/assignments/template4.qmd) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 The following will set-up [`markmyassignment`](https://github.com/MansMeg/markmyassignment) to check your functions at the end of the notebook:
- 
-```{r} 
+
+```{r}
 if(!require(markmyassignment)){
     install.packages("markmyassignment")
     library(markmyassignment)
 }
 assignment_path = paste("https://github.com/avehtari/BDA_course_Aalto/",
 "blob/master/assignments/tests/assignment4.yml", sep="")
-set_assignment(assignment_path)    
-```       
+set_assignment(assignment_path)
+```
 
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -68,7 +67,7 @@ if(!require(latex2exp)){
 ::::
 
 
-# Bioassay model 
+# Bioassay model
 
 ## (a)
 
@@ -165,15 +164,14 @@ posterior_mean <- function(alpha, beta) {
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 The following will check the functions for which `markmyassignment` has been set up:
- 
-```{r}  
-mark_my_assignment()    
-```      
+
+```{r}
+mark_my_assignment()
+```
 
 :::
 ::::
-
diff --git a/assignments/template5.qmd b/assignments/template5.qmd
index f1a7d4c0..a039645b 100644
--- a/assignments/template5.qmd
+++ b/assignments/template5.qmd
@@ -1,15 +1,15 @@
 ---
 title: "Assignment 5"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
@@ -22,37 +22,36 @@ editor: source
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
 This is the template for [assignment 5](assignment5.html). You can download the [qmd-file](https://avehtari.github.io/BDA_course_Aalto/assignments/template5.qmd) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 The following will set-up [`markmyassignment`](https://github.com/MansMeg/markmyassignment) to check your functions at the end of the notebook:
- 
-```{r} 
+
+```{r}
 if(!require(markmyassignment)){
     install.packages("markmyassignment")
     library(markmyassignment)
 }
 assignment_path = paste("https://github.com/avehtari/BDA_course_Aalto/",
 "blob/master/assignments/tests/assignment5.yml", sep="")
-set_assignment(assignment_path)    
-```       
+set_assignment(assignment_path)
+```
 
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -112,19 +111,19 @@ density_ratio <- function(alpha_propose, alpha_previous, beta_propose, beta_prev
     # beta_propose = 20.04, beta_previous = 24.76,
     # x = bioassay$x, y = bioassay$y, n = bioassay$n
     0.7661784
-    
+
 }
 # Then implement a function called `metropolis_bioassay()` which
 # implements the Metropolis algorithm using the `density_ratio()`:
 metropolis_bioassay <- function(alpha_initial, beta_initial, alpha_sigma, beta_sigma, no_draws, x, y, n){
     # Do computation here, and return as below.
-    # Below are "wrong" values (unlikely to actually occur) 
+    # Below are "wrong" values (unlikely to actually occur)
     # in the "correct" format (such that they work with the plotting functions further down).
     data.frame(
-        alpha=c(alpha_initial, alpha_initial+alpha_sigma, alpha_initial-alpha_sigma), 
+        alpha=c(alpha_initial, alpha_initial+alpha_sigma, alpha_initial-alpha_sigma),
         beta=c(beta_initial, beta_initial+beta_sigma, beta_initial-beta_sigma)
     )
-    
+
 }
 df = metropolis_bioassay(0, 0, 1, 1, 1000, bioassay$x, bioassay$y, bioassay$n)
 ```
@@ -177,15 +176,14 @@ mcmc_scatter(df, pars=c("alpha", "beta"))
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 The following will check the functions for which `markmyassignment` has been set up:
- 
-```{r}  
-mark_my_assignment()    
-```      
+
+```{r}
+mark_my_assignment()
+```
 
 :::
 ::::
-
diff --git a/assignments/template6.qmd b/assignments/template6.qmd
index 440585bc..16283de5 100644
--- a/assignments/template6.qmd
+++ b/assignments/template6.qmd
@@ -1,15 +1,15 @@
 ---
 title: "Assignment 6"
 author: anonymous # <-- hand in anonymously
-format: 
+format:
   html:
     toc: true
     code-tools: true
-    code-line-numbers: true  
+    code-line-numbers: true
     number-sections: true
     mainfont: Georgia, serif
     page-layout: article
-  pdf:  
+  pdf:
     geometry:
     - left=1cm,top=1cm,bottom=1cm,right=7cm
     number-sections: true
@@ -22,19 +22,19 @@ editor: source
 
 This is the template for [assignment 6](assignment6.html). You can download the [broken stan-file](./additional_files/assignment6_linear_model.stan) and the [qmd-file](https://avehtari.github.io/BDA_course_Aalto/assignments/template6.qmd) or copy the code from this rendered document after clicking on `</> Code` in the top right corner.
 
-**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.** 
+**Please replace the instructions in this template by your own text, explaining what you are doing in each exercise.**
 
 
 
 :::: {.content-hidden when-format="pdf"}
 ::: {.callout-warning collapse=false}
- 
-## Setup 
+
+## Setup
 
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
@@ -43,8 +43,7 @@ JupyterHub has all the needed packages pre-installed.
 The following installs and loads the `aaltobda` package:
 ```{r}
 if(!require(aaltobda)){
-    install.packages("remotes")
-    remotes::install_github("avehtari/BDA_course_Aalto", subdir = "rpackage", upgrade="never")
+    install.packages("aaltobda", repos = c("https://avehtari.github.io/BDA_course_Aalto/", getOption("repos")))
     library(aaltobda)
 }
 ```
@@ -121,7 +120,7 @@ Write your answers/code here!
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
@@ -130,7 +129,7 @@ Write your answers/code here!
 #| warning: false
 # These are our observations y: the proportion of students handing in each assignment (1-8),
 # sorted by year (row-wise) and assignment (column-wise).
-# While the code suggest a matrix structure, 
+# While the code suggest a matrix structure,
 # the result will actually be a vector of length N = no_years * no_assignments
 propstudents<-c(c(176, 174, 158, 135, 138, 129, 126, 123)/176,
                 c(242, 212, 184, 177, 174, 172, 163, 156)/242,
@@ -140,7 +139,7 @@ propstudents<-c(c(176, 174, 158, 135, 138, 129, 126, 123)/176,
 # These are our predictors x: for each observation, the corresponding assignment number.
 assignment <- rep(1:8, 5)
 # These are in some sense our test data: the proportion of students handing in the last assignment (9),
-# sorted by year. 
+# sorted by year.
 # Usually, we would not want to split our data like that and instead
 # use e.g. Leave-One-Out Cross-Validation (LOO-CV, see e.g. http://mc-stan.org/loo/index.html)
 # to evaluate model performance.
@@ -158,8 +157,8 @@ model_data = list(N=length(assignment),
 ```
 **Sampling from the posterior distribution happens here**:
 ```{r}
-#| warning: false 
-# This reads the file at the specified path and tries to compile it. 
+#| warning: false
+# This reads the file at the specified path and tries to compile it.
 # If it fails, an error is thrown.
 retention_model = cmdstan_model("./additional_files/assignment6_linear_model.stan")
 # This "out <- capture.output(...)" construction suppresses output from cmdstanr
@@ -175,21 +174,21 @@ out <- capture.output(
 # This extracts the draws from the sampling result as a data.frame.
 draws_df = fit$draws(format="draws_df")
 
-# This does some data/draws wrangling to compute the 5, 50 and 95 percentiles of 
-# the mean at the specified covariate values (x_predictions). 
+# This does some data/draws wrangling to compute the 5, 50 and 95 percentiles of
+# the mean at the specified covariate values (x_predictions).
 # It can be instructive to play around with each of the data processing steps
 # to find out what each step does, e.g. by removing parts from the back like "|>  gather(pct,y,-x)"
 # and printing the resulting data.frame.
-mu_quantiles_df = draws_df |> 
-      subset_draws(variable = c("mu_pred")) |> 
-      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |> 
-      mutate(x = 1:9) |> 
+mu_quantiles_df = draws_df |>
+      subset_draws(variable = c("mu_pred")) |>
+      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |>
+      mutate(x = 1:9) |>
       pivot_longer(c(q5, q50, q95), names_to = c("pct"))
 # Same as above, but for the predictions.
-y_quantiles_df = draws_df |> 
-      subset_draws(variable = c("y_pred")) |> 
-      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |> 
-      mutate(x = 1:9) |> 
+y_quantiles_df = draws_df |>
+      subset_draws(variable = c("y_pred")) |>
+      summarise_draws(~quantile2(.x, probs = c(0.05, .5, 0.95))) |>
+      mutate(x = 1:9) |>
       pivot_longer(c(q5, q50, q95), names_to = c("pct"))
 ```
 
@@ -203,14 +202,14 @@ y_quantiles_df = draws_df |>
 #| label: fig-posterior
 #| fig-cap: Describe me in your submission!
 ggplot() +
-  # scatter plot of the training data:  
+  # scatter plot of the training data:
   geom_point(
-    aes(x, y, color=assignment), 
+    aes(x, y, color=assignment),
     data=data.frame(x=assignment, y=propstudents, assignment="1-8")
 ) +
   # scatter plot of the test data:
   geom_point(
-    aes(x, y, color=assignment), 
+    aes(x, y, color=assignment),
     data=data.frame(x=no_assignments, y=propstudents9, assignment="9")
 ) +
   # you have to tell us what this plots:
@@ -239,7 +238,7 @@ ggplot() +
 
 *This block will only be visible in your HTML output, but will be hidden when rendering to PDF with quarto for the submission.*
 **Make sure that this does not get displayed in the PDF!**
-    
+
 
 
 
@@ -261,7 +260,7 @@ fit$cmdstan_diagnose()
 Write your answers/code here!
 
 
-# Generalized linear model: Bioassay with Stan (4 points) 
+# Generalized linear model: Bioassay with Stan (4 points)
 
 ## (d)
 
@@ -291,4 +290,3 @@ Write your answers/code here!
 
 
 Write your answers/code here!
-