search

memphis-iis / datawhys-content-notebooks-python

Content for DataWhys in the form of JupyterLab notebooks (.ipynb files)
Apache License 2.0
8 stars 2 forks source link

Clustering dendrogram y-axis off #99

Closed aolney closed 3 years ago

aolney commented 3 years ago

2020-09-22_08-36

aolney commented 3 years ago

Dale do you want to take a crack since you made the images? Perhaps it could be addressed with a caption or something instead?

ddbowman commented 3 years ago

Hi Andrew, I am getting an error when I try to respond on github. This example is from our multivariate textbook, "methods of multivariate analysis" by Rencher and Christensen. In their dendrogram plot the axis goes from 0 to 700+ and so it makes more sense to compare the numbers directly with the distances. I created the dendrogram using a package in R and it cut off the values between 0 and 350 which is misleading. I wonder if you could add a comment to the effect that the graphs is for illustration only to see which groups were combined at each level. Best, Dale

From: Andrew M Olney @.> Sent: Monday, May 31, 2021 11:53 AM To: memphis-iis/datawhys-content-notebooks @.> Cc: Dale D Bowman (ddbowman) @.>; Assign @.> Subject: Re: [memphis-iis/datawhys-content-notebooks] Clustering dendrogram y-axis off (#99)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and trust the content is safe.

Dale do you want to take a crack since you made the images? Perhaps it could be addressed with a caption or something instead?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHubhttps://github.com/memphis-iis/datawhys-content-notebooks/issues/99#issuecomment-851596532, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AOUET7W73WHVK6XXSUT24HLTQO5KFANCNFSM4RVZGZ4Q.

aolney commented 3 years ago

@ddbowman That would work; if regenerating the figure isn't easy, what do you think about me editing the image so the top horizontal line is at 590 (approximately marked in green below)? I think that's what the student was asking (the comment was from the online course)

image

ddbowman commented 3 years ago

Hi Andrew, That sounds fine to me. The dendrogram program I used was not very flexible. I think a commment that the y-axis might be misleading as far as the actual distances might be appropriate as well. Best, Dale

From: Andrew M Olney @.> Sent: Monday, May 31, 2021 2:40 PM To: memphis-iis/datawhys-content-notebooks @.> Cc: Dale D Bowman (ddbowman) @.>; Mention @.> Subject: Re: [memphis-iis/datawhys-content-notebooks] Clustering dendrogram y-axis off (#99)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and trust the content is safe.

@ddbowmanhttps://github.com/ddbowman That would work; if regenerating the figure isn't easy, what do you think about me editing the image so the top horizontal line is at 590 (approximately marked in green below)? I think that's what the student was asking (the comment was from the online course)

[image]https://user-images.githubusercontent.com/4432084/120234626-03978080-c21e-11eb-93ab-c1b807cfc963.png

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/memphis-iis/datawhys-content-notebooks/issues/99#issuecomment-851650069, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AOUET7VVMRSVJ32RYDXTUBDTQPQ2PANCNFSM4RVZGZ4Q.

aolney commented 3 years ago

Or, I could change the y axis to go from 0 to 700, with 590 lining up with the original horizontal line - that seems like the best if you think it would no longer be misleading

ddbowman commented 3 years ago

That would be optimal, I think. Best, Dale

From: Andrew M Olney @.> Sent: Monday, May 31, 2021 5:27 PM To: memphis-iis/datawhys-content-notebooks @.> Cc: Dale D Bowman (ddbowman) @.>; Mention @.> Subject: Re: [memphis-iis/datawhys-content-notebooks] Clustering dendrogram y-axis off (#99)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and trust the content is safe.

Or, I could change the y axis to go from 0 to 700, with 590 lining up with the original horizontal line - that seems like the best if you think it would no longer be misleading

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/memphis-iis/datawhys-content-notebooks/issues/99#issuecomment-851703736, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AOUET7THKHENKCVNU6ILNWLTQQEMHANCNFSM4RVZGZ4Q.

aolney commented 3 years ago

Ended up implementing the spirit of this, but not with 0 and 700 axis labels, because the figure wouldn't support it.

However the new labels are consistent with the distances in the preceding tables

Closing with https://github.com/memphis-iis/datawhys-content-notebooks/commit/f5f84bedb6ded11d96130a8784e82d1e0bee8787