[BioC] problem with MANOR

[J.delasHeras at ed.ac.uk]

Hi Michela,

the warnings are pointing at thye problem, and the solution: "make  
'span' larger". You specified a span or 0.03, so only 3% of the spots  
are used to infer the trend, and at least in some areas (as  
indicated), there are too few spots, so the warning suggests you  
increase the 'span' parameter to get a better estimate.
Try span=0.3, or higher if you need. The max is span=1, where *all*  
the spots are considered.

Jose

Quoting Ballardini Michela <m.ballardini at ior-forli.it>:

> Hello,
>
> I have a problem with the package MANOR.
>
>
>
> I import a ‘default’ file and I use R 2.5.0 version.
>
> This is the output when I try to estimate the spatial trend:
>
>
>
>> arrayABTrend <- arrayTrend(arrayAB, "LogRatio", span=0.03, degree=1,
>
> + iterations=3, family="symmetric")
>
> Errore: NA/NaN/Inf in chiamata a funzione esterna (arg 1)
>
> Inoltre: Ci sono 50 o più avvisi (usare warnings() per leggere i primi 50)
>
>>
>
>> warnings()
>
> Messaggi di avvertimento:
>
> 1: at  0.98503 0.51586
>
> 2: radius  0.00057507
>
> 3: all data on boundary of neighborhood. make span bigger
>
> 4: pseudoinverse used at 0.98503 0.51586
>
> 5: neighborhood radius 0.023981
>
> 6: reciprocal condition number  -1.#IND
>
> 7: at  4.0151 0.51586
>
> 8: radius  0.00057507
>
> 9: all data on boundary of neighborhood. make span bigger
>
> 10: There are other near singularities as well. 0.00057507
>
> 11: at  0.98503 4.2953
>
> 12: radius  0.00057507
>
> 13: all data on boundary of neighborhood. make span bigger
>
> 14: at  4.0151 4.2953
>
> 15: radius  0.00057507
>
> 16: all data on boundary of neighborhood. make span bigger
>
> 17: at  0.98503 2.1383
>
> 18: radius  0.00022501
>
> 19: all data on boundary of neighborhood. make span bigger
>
> 20: at  4.0151 2.1383
>
> 21: radius  0.00022501
>
> 22: all data on boundary of neighborhood. make span bigger
>
> 23: at  3.0001 0.51586
>
> 24: radius  0.00035006
>
> 25: all data on boundary of neighborhood. make span bigger
>
> 26: zero-width neighborhood. make span bigger
>
> 27: zero-width neighborhood. make span bigger
>
> 28: at  2.0001 4.2953
>
> 29: radius  0.00035006
>
> 30: all data on boundary of neighborhood. make span bigger
>
> 31: at  2.0001 0.51586
>
> 32: radius  0.00035006
>
> 33: all data on boundary of neighborhood. make span bigger
>
> 34: zero-width neighborhood. make span bigger
>
> 35: at  4.0151 1.0691
>
> 36: radius  0.00022501
>
> 37: all data on boundary of neighborhood. make span bigger
>
> 38: at  0.98503 3.2074
>
> 39: radius  0.00022501
>
> 40: all data on boundary of neighborhood. make span bigger
>
> 41: zero-width neighborhood. make span bigger
>
> 42: at  4.0151 3.2074
>
> 43: radius  0.00022501
>
> 44: all data on boundary of neighborhood. make span bigger
>
> 45: at  0.98503 1.0691
>
> 46: radius  0.00022501
>
> 47: all data on boundary of neighborhood. make span bigger
>
> 48: zero-width neighborhood. make span bigger
>
> 49: at  4.0001 0.51586
>
> 50: radius  0.00035006
>
>
>
> Can you explain me what can I do to resolve the problem, please?
>
>
>
> Michela
>
>
>
> *************************************************
>
> Michela Ballardini
>
> Unità di Biostatistica e Sperimentazioni Cliniche - Pad. Valsalva
>
> c/o OSP. MORGAGNI-PIERANTONI
>
> Via Forlanini, 34
>
> 47100 Forlì
>
> Tel. (+39) 0543.731836
>
> Fax (+39) 0543.731612
>
> *************************************************
>
>
>
>
> 	[[alternative HTML version deleted]]
>
>



-- 
Dr. Jose I. de las Heras                      Email: J.delasHeras at ed.ac.uk
The Wellcome Trust Centre for Cell Biology    Phone: +44 (0)131 6513374
Institute for Cell & Molecular Biology        Fax:   +44 (0)131 6507360
Swann Building, Mayfield Road
University of Edinburgh
Edinburgh EH9 3JR
UK