3-D Scatter (XYZ) Chart in Office 2003?

  • Thread starter Thread starter Guest
  • Start date Start date
Andy, Thanks for the tip ... Fernando's four quadrant scatter diagram is
exactly what I'm looking for. Unfortunatly, I don't speak the languange! Can
you by any chance explain how to use the add-in?

Thanks!
 
Hi,

Are you are referring to Fernando's 3d.xls sample file. Unfortunately
this is protected. You can plot up to 200 points by simply placing your
xyz values in the available cells.

If that's not the example you are referring to post back.

Cheers
Andy
 
I've downloaded the "Cluster.exe" add-in. He's provided the password to use
the it, but I'm prompted to enter axis criteria and I can't understand the
language. I'll try placing my data in his spreadsheet and see if I can work
it that way.
 
It is the least one can do to promote the brilliant work of Mr. Cinquegrani
to provide this brief translation:

In the esempioCluster.xls workbook (esempio=example; it has just one
worksheet):
-- column headings "Primo asse" "Secondo asse" ... = First axis, Second axis
....
-- text below:
Aggregazione in gruppi e dendrogramma = aggregation into groups and
dendrogram

suggerimenti di Jon Peltier e John Walkenbach = suggestions by J.P. & J.W.

avviare cluster.xla = launch ...

richiamare da menu Cluster/Password e immettere la password = from the
Cluster menu, select Password and enter ...

selezionare la zona in giallo = select the area (range) in yellow

avviare da menu Cluster/Cluster = from the Cluster menu, select Cluster



In the clustering dialog:

Individui = Individuals (case labels in usual statistical terminology)
[identified as A2:A16 if you follow the example]

Variabili = Variables (variable labels) [B1:F1]

Dati = Data [B2:F16]

Cluster slider = No.of clusters (available 1..12) [default 5]



<<the add-in does hierarchical clustering, but the pre-specified number of
clusters is used for visualistion with convex hulls -- something I did some
work on -- anyone interested see the R package chplot and the corresponding
article in CMPB 2005 vol. 78 no.1 p.69-74)>>



Legame = Linkage

-- singolo = single

-- completo = complete

-- medio = average

-- medio ponderato = weighted average

<<OTOMH, I can't tell which of the many clustering algorithms the last one
is -- I hope it's Wards, since about a zillion studies have shown that one
should just about forget every other algorithm in clustering, and, BTW, any
measure other than Euclidean distance (unless the data are polytomous or
compositional)>>



Grafico XY = XY Graphics

-- etichette = labels

-- inviluppi = envelopes (in this context it means convex hulls)



Dendrogramma = Dendrogram

-- normale = normal

-- logaritmico = logarithmic

-- quadratico = quadratic

<<this regulates the scaling of the distance axis>>



Metrica = Metrics (i.e., distance)

-- Euclidea = Eculidean

-- Euclidea^2 = Squared Eculidean

(the remaining two do not need translation)



Buttons:

anulla = cancel

Esegui = Execute (Run)



<<Ecco, fatta la traduzione che, per dir' la verita, non era ne anche troppo
necessaria, perche quasi tutte le parole (questo commento incluso) sono di
origine Latina, quindi per uno che parla l' Inglese e si intende di
clustering e di Excel lo significato dovrebbe essere ovvio.>>



Regards with apologies for occasional typos and misspellings,



Gaj Vidmar

Univ. of Ljubljana, Fac. of Medicine, Inst. of Biomedical Informatics

Ljubljana, Slovenia, EU, Europe
 
Thanks so much for your help! I'll try this out and post back after I've
attempted it.

Thanks again. You're wonderful!

Gaj Vidmar said:
It is the least one can do to promote the brilliant work of Mr. Cinquegrani
to provide this brief translation:

In the esempioCluster.xls workbook (esempio=example; it has just one
worksheet):
-- column headings "Primo asse" "Secondo asse" ... = First axis, Second axis
....
-- text below:
Aggregazione in gruppi e dendrogramma = aggregation into groups and
dendrogram

suggerimenti di Jon Peltier e John Walkenbach = suggestions by J.P. & J.W.

avviare cluster.xla = launch ...

richiamare da menu Cluster/Password e immettere la password = from the
Cluster menu, select Password and enter ...

selezionare la zona in giallo = select the area (range) in yellow

avviare da menu Cluster/Cluster = from the Cluster menu, select Cluster



In the clustering dialog:

Individui = Individuals (case labels in usual statistical terminology)
[identified as A2:A16 if you follow the example]

Variabili = Variables (variable labels) [B1:F1]

Dati = Data [B2:F16]

Cluster slider = No.of clusters (available 1..12) [default 5]



<<the add-in does hierarchical clustering, but the pre-specified number of
clusters is used for visualistion with convex hulls -- something I did some
work on -- anyone interested see the R package chplot and the corresponding
article in CMPB 2005 vol. 78 no.1 p.69-74)>>



Legame = Linkage

-- singolo = single

-- completo = complete

-- medio = average

-- medio ponderato = weighted average

<<OTOMH, I can't tell which of the many clustering algorithms the last one
is -- I hope it's Wards, since about a zillion studies have shown that one
should just about forget every other algorithm in clustering, and, BTW, any
measure other than Euclidean distance (unless the data are polytomous or
compositional)>>



Grafico XY = XY Graphics

-- etichette = labels

-- inviluppi = envelopes (in this context it means convex hulls)



Dendrogramma = Dendrogram

-- normale = normal

-- logaritmico = logarithmic

-- quadratico = quadratic

<<this regulates the scaling of the distance axis>>



Metrica = Metrics (i.e., distance)

-- Euclidea = Eculidean

-- Euclidea^2 = Squared Eculidean

(the remaining two do not need translation)



Buttons:

anulla = cancel

Esegui = Execute (Run)



<<Ecco, fatta la traduzione che, per dir' la verita, non era ne anche troppo
necessaria, perche quasi tutte le parole (questo commento incluso) sono di
origine Latina, quindi per uno che parla l' Inglese e si intende di
clustering e di Excel lo significato dovrebbe essere ovvio.>>



Regards with apologies for occasional typos and misspellings,



Gaj Vidmar

Univ. of Ljubljana, Fac. of Medicine, Inst. of Biomedical Informatics

Ljubljana, Slovenia, EU, Europe
Click to expand...
 
Back
Top