[Mondrian] RES: RES: Percentile() function

Kleyson Rios kleyson.rios at saude.go.gov.br
Thu Aug 8 09:37:19 EDT 2013


Paul,

 

Based on JIRA MONDRIAN-1354, the fix will be available only in
<http://jira.pentaho.com/browse/MONDRIAN/fixforversion/10716> 3.6.0 (5.0.0
GA Suite Release),
<http://jira.pentaho.com/browse/MONDRIAN/fixforversion/11537> Mondrian 4.0
Sprint 3, and I am using Mondrian-3.5.7

 

I also realized a problem with ThirdQ() function.

 

For the following query :

 

with

 

member [Measures].[Median] as 'Median([Ano].[Ano].Members,
[Measures].[Incidencia])'

member [Measures].[ThirdQ] as 'ThirdQ([Ano].[Ano].Members,
[Measures].[Incidencia])'

 

select 

 

NON EMPTY {[Measures].[Median], [Measures].[ThirdQ]} ON COLUMNS,

NON EMPTY {[Semana].[Semana].Members} ON ROWS

 

from [IA Dengue]

 

where {[Municipio].[Goiânia]}

 

The result is:

 

(
)

Semana: 3

Median: 19,793562 (correct)

ThirdQ: 33,987053 (correct)

(..)

 

 

But this other query:

 

with

 

member [Measures].[Median] as 'Median([Ano].[Ano].Members,
[Measures].[Incidencia])'

member [Measures].[ThirdQ] as 'ThirdQ([Ano].[Ano].Members,
[Measures].[Incidencia])'

 

select 

 

NON EMPTY {[Measures].[Median], [Measures].[ThirdQ]} ON COLUMNS,

NON EMPTY {[Semana].[Semana].Members} ON ROWS

 

from [IA Dengue]

 

where Crossjoin({[Epidemico].[N]}, {[Municipio].[Goiânia]})

 

 

The result is:

 

(
)

Semana: 3

Median: 14,344192 (correct)

ThirdQ: 29,017727 (wrong. Based on Excel and R should be 26,7116858)

(..)

 

For all ‘Semanas’,  Median() is correct but for ThirdQ() is wrong.

 

I tried to reproduce on SteelWheel but without success.

 

 

Dataset for the FIRST query (2.235699, 7.556526, 35.50954, 13.262299,
10.411041, 5.495393, 50.310879, 15.426085, 20.230853, 29.017727, 291.575645,
33.987053, 19.793562)

 

Dataset for the SECOND query (2.235699, 7.556526, 13.262299, 10.411041,
5.495393, 50.310879, 15.426085, 29.017727, 33.987053, 19.793562)

 

 

Regards.

 

-- 

Kleyson Rios

Gestor de TI

Coordenador de BI / Banco de Dados

Gerência de Tecnologia da Informação

Secretaria de Estado da Saúde de Goiás

(62) 3201-3707

 

Se não pode vencer pelo talento, vença pelo esforço. (Dave Weinbaum)

 

De: mondrian-bounces at pentaho.org [mailto:mondrian-bounces at pentaho.org] Em
nome de Paul Stoellberger
Enviada em: quinta-feira, 8 de agosto de 2013 09:54
Para: Mondrian developer mailing list
Assunto: Re: [Mondrian] RES: Percentile() function

 

Yeah, I fixed that as well

 

https://github.com/pentaho/mondrian/pull/44

 

Kleyson, this version should include the fix so I don't know why you're
getting wrong results for  Percentile75

Can you retest with a trunk snapshot build by any chance?

 

-Paul

 

On Aug 8, 2013, at 2:44 PM, Brandon Jackson wrote:





About the JIRA did the median of a single element case get fixed as well?

I'll double check if the service packs included this in 4.8.2 because we
actually need to use this function next week.

Brandon

Sent from my iPhone

On Aug 7, 2013, at 12:34 PM, Julian Hyde <jhyde at pentaho.com> wrote:




On Aug 7, 2013, at 3:20 AM, Paul Stoellberger <p.stoellberger at gmail.com>
wrote:

 

What version of mondrian are you using?

There used to be a bug that would return negative values in some cases

 

The bug was MONDRIAN-1354.

 

Julian

 

 

_______________________________________________

Mondrian mailing list

Mondrian at pentaho.org

http://lists.pentaho.org/mailman/listinfo/mondrian

_______________________________________________
Mondrian mailing list
Mondrian at pentaho.org
http://lists.pentaho.org/mailman/listinfo/mondrian

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20130808/2043bff5/attachment.html 


More information about the Mondrian mailing list