[Gvsig_english] gvSIG and Java on Mac OS X

[Agustin Diez Castillo]

Ben,
I was on the field. I have checked then with several commands. Example  
nearest neighbour:
# your bundle:
Vecino más cercano
Distancia media:0
Distancia esperada:7,009
Distancia esperada corregida con efecto borde:7,068
Varianza:0,005
Varianza corregida con efecto borde:0,006
Índice de vecino mas próximo:0
Índice de vecino mas próximo corregido:0
Valor Z:-94,576
Valor Z corregido:-92,075
# The same shape with sextante 0.53
          Distancia media:0,911
Distancia esperada:7,009
Distancia esperada corregida con efecto borde:7,068
Varianza:0,005
Varianza corregida con efecto borde:0,006
Índice de vecino mas próximo:0,13
Índice de vecino mas próximo corregido:0,129
Valor Z:-82,278
Valor Z corregido:-80,202
It's obvious that yours (maybe when running in a Spanish locale) lost  
the decimals. I have checked it with several set of points and it  
happens consistently HOWEVER quadrant analysis produces the very same  
results. May first guess was that maybe there was some problem with  
big set of points but the same happens with smaller sets (see the last  
example)
#Yours
Análisis por cuadrantes
Media1,996732
Varianza =39,511427
Coeficiente de variación [~0 = uniforme, ~1 = aleatorio, >1 =  
agregado] =19,788047
Valor D del Test de Kolmogorov-Smirnov =0,731052
Valor crítico al 5% =0,038873
#sextante 0.53

        nálisis por cuadrantes

Media1,996732
Varianza =39,511427
Coeficiente de variación [~0 = uniforme, ~1 = aleatorio, >1 =  
agregado] =19,788047
Valor D del Test de Kolmogorov-Smirnov =0,731052
Valor crítico al 5% =0,038873

With the attached set:
Vecino más cercano
Distancia media:0
Distancia esperada:2,482
Distancia esperada corregida con efecto borde:2,56
Varianza:0,009
Varianza corregida con efecto borde:0,01
Índice de vecino mas próximo:0
Índice de vecino mas próximo corregido:0
Valor Z:-26,577
Valor Z corregido:-25,166
        Vecino más cercano

Distancia media:2,239
Distancia esperada:2,482
Distancia esperada corregida con efecto borde:2,56
Varianza:0,009
Varianza corregida con efecto borde:0,01
Índice de vecino mas próximo:0,902
Índice de vecino mas próximo corregido:0,875
Valor Z:-2,593
Valor Z corregido:-3,154
El Oct 20, 2008, a las 7:45 PM, Benjamin Ducke escribió:

> Agustin,
>
> which command exactly was it and what parameter settings
> did you use?
>
> If you can send me the original input data, I can try to
> reproduce this with different SEXTANTE versions on Windows
> and Linux. This way, we might find out where the problem
> originates.
>
> You are saying "Intel works, but not PPC". Does that mean
> you get consistently the same results on an Intel CPU
> but not a PPC? If so, there might be a problem with the
> different floating point representations.
>
> Ben
>
> ----- Original Message -----
> From: "Agustin Diez Castillo" <Agustin.Diez at uv.es>
> To: "Users and Developers mailing list" <gvsig_internacional at runas.cap.gva.es 
> >
> Sent: Monday, October 20, 2008 2:12:45 PM (GMT) Europe/London
> Subject: Re: [Gvsig_english] gvSIG and Java on Mac OS X
>
> Ben,
> Neither do I. I will double check and report to victor. It's true that
> I'm using a different version (former 0.53v).
> Anyhow,
> I got very different results with exactly the same command, not even
> the number of valid cells is the same, weird.
> It's working on 32 bit intel but not in ppcs.
> Agustin
>
>
>
> ------
> Files attached to this email may be in ISO 26300 format (OASIS Open  
> Document Format). If you have difficulty opening them, please visit http://iso26300.info 
>  for more information.
>
> _______________________________________________
> Gvsig_internacional mailing list
> Gvsig_internacional at runas.cap.gva.es
> http://runas.cap.gva.es/mailman/listinfo/gvsig_internacional
>

-------------- next part --------------
Skipped content of type multipart/mixed