'help'

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

'help'

Yalemzewod Gelaw
Hi

I tried to compute the Moran's I statistics using GIS and R. However, I got
different values.

Could you please brief me how the difference can occurs?

Using GIS

 Global Moran's I Summary

Moran's Index:   0.151485

Expected Index:  -0.007692

Variance:        0.001040

z-score:         4.935027

p-value:         0.000001



Using R

Moran I test under randomisation

data:  Dat$allprop

weights: Xlist

Moran I statistic standard deviate = 0.82453, p-value = 0.4096

alternative hypothesis: two.sided

sample estimates:

Moran I statistic       Expectation          Variance

      0.034154753      -0.007299270       0.002527644

NB: in the neighbour links there are 2 regions with no links


Thank you for any help



*Regards, *

Yalem

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: 'help'

Sarah Goslee
Hi,

It's pretty much impossible, since we don't know what GIS software you
used, what commands, what R package and code, or what your data look
like.

As a starting point, I'd suggest carefully rereading the help files
for both the GIS and the R functions, to make sure that they are doing
the same thing, and what you expect them to, and to review your data
in both places to ensure that it is the same values and projection,
etc. It is entirely possible that one software is scaling your data
for you automatically, or some other processing step that is not
duplicated by the other software.

If trying those things doesn't clarify it for you, then please post
back to the list with more information about the software and data
you're using.

Sarah

On Sat, Apr 14, 2018 at 1:18 AM, Yalemzewod Gelaw <[hidden email]> wrote:

> Hi
>
> I tried to compute the Moran's I statistics using GIS and R. However, I got
> different values.
>
> Could you please brief me how the difference can occurs?
>
> Using GIS
>
>  Global Moran's I Summary
>
> Moran's Index:   0.151485
>
> Expected Index:  -0.007692
>
> Variance:        0.001040
>
> z-score:         4.935027
>
> p-value:         0.000001
>
>
>
> Using R
>
> Moran I test under randomisation
>
> data:  Dat$allprop
>
> weights: Xlist
>
> Moran I statistic standard deviate = 0.82453, p-value = 0.4096
>
> alternative hypothesis: two.sided
>
> sample estimates:
>
> Moran I statistic       Expectation          Variance
>
>       0.034154753      -0.007299270       0.002527644
>
> NB: in the neighbour links there are 2 regions with no links
>
>
> Thank you for any help
>
>
>
> *Regards, *
>
> Yalem



--
Sarah Goslee
http://www.functionaldiversity.org

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: 'help'

Roger Bivand
Administrator
On Sat, 14 Apr 2018, Sarah Goslee wrote:

> Hi,
>
> It's pretty much impossible, since we don't know what GIS software you
> used, what commands, what R package and code, or what your data look
> like.
>
> As a starting point, I'd suggest carefully rereading the help files
> for both the GIS and the R functions, to make sure that they are doing
> the same thing, and what you expect them to, and to review your data
> in both places to ensure that it is the same values and projection,
> etc. It is entirely possible that one software is scaling your data
> for you automatically, or some other processing step that is not
> duplicated by the other software.
>
> If trying those things doesn't clarify it for you, then please post
> back to the list with more information about the software and data
> you're using.

Yes, the problem is most likely that the weights and data are the same,
but in GIS or R the data are sorted differently than the weights.
Alternatively, the weights may be different for the same data. For R, it
would help to see how you constructed the weights object. If the GIS is
ArcGIS, for the same weights and data, we know what differences may be
expected, and any such differences would not affect your conclusion.

Roger

>
> Sarah
>
> On Sat, Apr 14, 2018 at 1:18 AM, Yalemzewod Gelaw <[hidden email]> wrote:
>> Hi
>>
>> I tried to compute the Moran's I statistics using GIS and R. However, I got
>> different values.
>>
>> Could you please brief me how the difference can occurs?
>>
>> Using GIS
>>
>>  Global Moran's I Summary
>>
>> Moran's Index:   0.151485
>>
>> Expected Index:  -0.007692
>>
>> Variance:        0.001040
>>
>> z-score:         4.935027
>>
>> p-value:         0.000001
>>
>>
>>
>> Using R
>>
>> Moran I test under randomisation
>>
>> data:  Dat$allprop
>>
>> weights: Xlist
>>
>> Moran I statistic standard deviate = 0.82453, p-value = 0.4096
>>
>> alternative hypothesis: two.sided
>>
>> sample estimates:
>>
>> Moran I statistic       Expectation          Variance
>>
>>       0.034154753      -0.007299270       0.002527644
>>
>> NB: in the neighbour links there are 2 regions with no links
>>
>>
>> Thank you for any help
>>
>>
>>
>> *Regards, *
>>
>> Yalem
>
>
>
>

--
Roger Bivand
Department of Economics, Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.
voice: +47 55 95 93 55; e-mail: [hidden email]
http://orcid.org/0000-0003-2392-6140
https://scholar.google.no/citations?user=AWeghB0AAAAJ&hl=en

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Roger Bivand
Department of Economics
Norwegian School of Economics
Helleveien 30
N-5045 Bergen, Norway