Confusion about HW 10.1 wording

[BoschiGiacomo]

I'm opening this issue because I find the wording of exercise 10.1 unclear, and i'm not sure of the intended solution. The text of the exercise states: "Compute the Euclidean distance matrix on the standardized data for the first three cities"

This phrasing seems to suggest that one should compute the standardized euclidean distances only on the first three cities. In the official solution, however, the distances are computed using the full dataset of cities, and only the first three results of the resulting distance matrix are displayed.

Since standardized Euclidean distances takes into account the context in which the variables are placed (e.g. it is different to consider two provinces in the context of northern Italy compared to all of Italy) these two procedures will generally yield different results.

Could you please clarify which interpretation is intended for this exercise?

If the goal is to standardize using all cities and compute the full distance matrix, perhaps the wording could be revised to something like:

“Compute the Euclidean distance matrix on the standardized data for all cities, and then show the distances for the first three cities in table format"

If instead the goal is to work only with the first three cities, the solution code may need to be updated accordingly.

Also, on a less important note, as this is the english version of the book, i suggest translating the variable name "nomiPrime3" to "namesFirst3"

Thank you in advance for your time.

[MarcoRianiUNIPR]

Dear Giacomo,
I tried to put the question below in ChatGPT to find out what it thinks

It says: if unclear, the standard default is: compute mean and SD using the full dataset.

This is exactly what we had assumed. However, you are right saying that we should have been more precise.

Please let me know what the community thinks. Was it obvious or not that the standardized data refer to the overall matrix?

[BoschiGiacomo]

I see, thanks for the clarification!