195 Reputation

5 Badges

16 years, 15 days

MaplePrimes Activity

These are replies submitted by Teep


Many thanks for this. Really interesting solution.

When I run this workbook on my machine (Maple 2018), I get an error message: Reserved word 'local' unexpected.

Perhaps this is related to different versions used?



Thanks so much ... a really nice routine ... the use of the sequence function is elegant.

Can you also suggest an approach to obtain the midpoint / centroid of each internal grid? 


@Carl Love 

Thank you. I'll try that approach ..

@Carl Love 

Each point is assigned a weight but the weight does not need to be unique to any point - in fact, thinking about this further, points may have common weights. I wanted to ensure that weights are not necessarily equal at each point in the scatterplot and so the question you asked is indeed valid. I apologise for not being specific here.


@Daniel Skoog 

This is useful - I appreciate the steer. This subject appears to be very interesting  - perhaps, fascinating ... I'm just beginning to touch the surface.

If I have the good fortune to advance this to the weighted cluster problem, I'll post my findings here.

You have my thanks. 




@Carl Love 

Good interpretation ... thanks.

I was thinkingof the weighted mean centre of a cluster - in any case, the distribution of each point weight will influence the location of the centre of mass in that region / cluster.

@Daniel Skoog @Carl Love

Extending the problem ...

Suppose each point is assigned a unique weight. Does anyone know whether this (or the fuzzy clustering) routine be modified to support weighted clustering? 


@Daniel Skoog 

That's a great help - I'm obliged.

@Daniel Skoog 


I (somehow) managed to download the file. It looks great! 

I'll work through the model and use it to help me understand this topic further.

Again - I really appreciate all the help and introduction to this subject.

@Daniel Skoog 

Just returning to this topic with a brief update.

Despite my lack of tech-savvy, I persevered and tried to download the ClusterAnalysis mla file from Github as you guggested. However, I receive the following error message: " ... we cannot show files that are this big right now". The size of the file appears to be 1.64MB and so, I assume this is too large to download (?). 

Following that, I am wondering if indeed it is possible to obtain access to the k-means worksheet?

Thanks for reading!

@Daniel Skoog@Carl Love

First-off, you're response is truly welcomed and appreciated ... I'm obliged, Daniel.

I did attempt to download the k-means package but alas, no luck (I use Maple 18, by the way).

I'm unsure as to how to download via Github (I'm not too tech-savy, I'm afraid), so is there any other means to download you can suggest? 

Reading the subject matter as Carl suggested, I believe what meets my needs is the k-means method, so I would love to have access to your routines.


Again ... my thanks to you.




Very interesting ... thanks.

The example given was illustrative only. I'll take a look at the source material you provided and explore further.

@Carl Love 

That's good ... hopefully this will help,


@Carl Love 

Thanks again, Carl.

Very interesting ... I am not familiar with this topic, so I really appreciate the steer.

I took your advice and trawled Mapleprimes and the Web but cannot see any light at the end of this tunnel ... now I'm even more intrigued!

If anyone can lend a hand here, it would be much appreciated.

(it appears that there is a Cluster Analysis package developed by Daniel Skoog, but I cannot seem to download this).



Yes - the data frame approach is used only to attach the labels to the Excel spreadsheet, but as you indicate - it may be redundant.

Your advice is warmly received and the additional insights are great to get. I'll use your routine in future. Thanks for enlightening me.

2 3 4 5 6 7 8 Page 4 of 9