8c. PCA of WHO SDGs and 13. Cluster Analysis

Task D

Author

AJ Smit

1 Task D

Question 1 refers to PCA SDG.

Explain the code section-by-section in long-form text. Include also the reasoning/rationale behind each section.

Questions 2–5 refer to Cluster Analysis.

What happens if we use pam() to create four, five, or even six clusters?
In your reasoned opinion, what would be the optimal number of clusters to use?
Repeat the analysis using either kmeans() or hclust(), and feel free to use the factoextra helper functions and visualisations. Are the results markedly different? Which clustering approach do you wish to proceed with — i.e., pam(), hclust() or kmeans()?
Describe the patterns that you observe at the end of your ordination and final cluster selection (i.e. based on the optimal number of clusters and whichever cluster technique you deem most appropriate). How does South Africa fare in terms of attaining SDGs? Contrast with some key countries of your choice to make your points. Label the key countries that you refer to in your text by updating the code accordingly. Continue to explain these patterns in terms of the global socio-political/socio-economic landscape. Provide a discourse about possible explanations for the patterns observed globally and regionally.

2 Assessment Criteria

This Task is not formally assessed.

Reuse

CC BY-NC-SA 4.0

Citation

BibTeX citation:

@online{smit,_a._j.,
  author = {Smit, A. J., and Smit, AJ},
  title = {8c. {PCA} of {WHO} {SDGs} and 13. {Cluster} {Analysis}},
  url = {http://tangledbank.netlify.app/BCB743/tasks/Task_D.html},
  langid = {en}
}

For attribution, please cite this work as:

Smit, A. J., Smit A 8c. PCA of WHO SDGs and 13. Cluster Analysis. http://tangledbank.netlify.app/BCB743/tasks/Task_D.html.

--- title: "8c. PCA of WHO SDGs and 13. Cluster Analysis" subtitle: "Task D" author: - "AJ Smit" --- ```{r code-brewing-opts, echo=FALSE} knitr::opts_chunk$set( comment = "R>", warning = FALSE, message = FALSE, fig.width = 4.5, fig.height = 2.625, out.width = "75%", fig.asp = NULL, # control via width/height dpi = 300 ) ggplot2::theme_set( ggplot2::theme_minimal(base_size = 8) ) ggplot2::theme_set( ggplot2::theme_bw(base_size = 8) ) ``` ## Task D **Question 1 refers to PCA SDG.** 1. Explain the code section-by-section in long-form text. Include also the reasoning/rationale behind each section. **Questions 2--5 refer to [Cluster Analysis](../cluster_analysis.qmd).** 2. What happens if we use `pam()` to create four, five, or even six clusters? 3. In your reasoned opinion, what would be the optimal number of clusters to use? 4. Repeat the analysis using either `kmeans()` or `hclust()`, and feel free to use the **factoextra** helper functions and visualisations. Are the results markedly different? Which clustering approach do you wish to proceed with --- i.e., `pam()`, `hclust()` or `kmeans()`? 5. Describe the patterns that you observe at the end of your ordination and final cluster selection (i.e. based on the optimal number of clusters and whichever cluster technique you deem most appropriate). How does South Africa fare in terms of attaining SDGs? Contrast with some key countries of your choice to make your points. Label the key countries that you refer to in your text by updating the code accordingly. Continue to explain these patterns in terms of the global socio-political/socio-economic landscape. Provide a discourse about possible explanations for the patterns observed globally and regionally.            ## Assessment Criteria This Task is not formally assessed.