(replying to the list)
Well, if you're willing to accept the indicator kriging values as
estimates of these probabilities, you're essentially done. Of course,
the kriged values can be outside [0, 1], so you need to deal with that.
Best regards,
--
Edzer
Zia Ahmed wrote:
> Dear Edzer,
>
> Thanks for your replay. I have done IK of water arsenic using
> quantiles of the data. I want to do miss classification of my risk
> that proposed by Goovaerts.
> Two misclassification of risks are proposed by Goovaerts (1997) based
> on ccdf model F(*u*;z_k |(n)) of Z(*u*)?Z_k .
>
>
> 1. The risk ?(*u*) of incorrectly classified a location u as
> hazardous (false positive):
>
> ?(*u*) = Prob{Z(*u*) ?z_c |z*_L (u) > Z_c , (n)}
>
> = F(*u*;z_c |(n))
>
> Where z*_L (u) is the estimator.
>
> 2. The risk ?(u) wrongly classified a location u as safe (false
> negative)
>
> ?(*u*) = Prob{Z(*u*) > z_c |z*_L (u) ? Z_c , (n)}
>
> = 1- F(*u*;z_c |(n))
>
>
>
>
>
> Ref: Goovaerts, P (1997). Geostatistics for Natural Resource
> Evaluation, pp: 259-368
>
>
>
> Zia
>
> Edzer Pebesma wrote:
>> Dear Zia,
>>
>> If you mean by Goovaerts (1997) his 483 page book "Geostatistics for
>> Natural Resource Evaluation", then please give us a bit more precise
>> detail on what you want (which page? which equation?)
>> --
>> Edzer
>>
>> Zia Ahmed wrote:
>>
>>> Hi all,
>>>
>>> Is it possible in gstat to get misclassification of risks as
>>> hazardous (false positive) and safe (false negative) as describe by
>>> Goovaerts (1997) from parametric or non-parametric probability
>>> mapping?. help will appreciated.
>>>
>>> Thanks
>>>
>>> Zia
>>>
>>>
>>>
>>
>>
>
--
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster
Weseler Straße 253, 48151 Münster, Germany. Phone: +49 251
8333081, Fax: +49 251 8339763 http://ifgi.uni-muenster.de/
http://www.springer.com/978-0-387-78170-9 e.pebe...@wwu.de
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
