[BioC] different probe level data at computer cluster
Martin Morgan
mtmorgan at fhcrc.org
Mon May 19 15:22:48 CEST 2008
Hi Markus --
pmindex eventually calls ls() on the cdf environment. ls() gets all
the probe names and then sorts them. This is where I bet your problem
occurs -- your master and worker nodes have different locales,
(de_DE.UTF-8 vs C) and so sort the probe names differently. Perhaps
you're getting the same answers, just in the different order? Try
sorting the results by probe name before all.equal.
Martin
Markus Schmidberger <schmidb at ibe.med.uni-muenchen.de> writes:
> Hello,
>
> to get the location of perfect matches from an AffyBatch I use the
> function pmindex().
> If I send my AffyBatch to a slave and use the pmindex() function at
> the slave I get an other result than using the function at the master
> node.
>
> Attached you can find some simple testcode and the sessionInfo plot.
> You can see, that the same annotation package will be used at the
> master and at the slaves. Therefore there should be the same index
> structure. But there is a difference!
>
> Any ideas for solutions?
>
> Best regards
> Markus
>
> ##################
> library(snow)
> c1<- makeCluster(1)
>
> library(affy)
> library(affydata)
> data(Dilution)
>
> indexMaster <- pmindex(Dilution)
> indexSlave <- clusterCall(c1, function(x) { library(affy);
> pmindex(x)}, Dilution) # Returns a LIST!
>
> all.equal(indexMaster, indexSlave[[1]])
> [1] "Names: 40 string mismatches" [2]
> "Component 12586: Numeric: lengths (69, 20) differ" [3] "Component
> 12587: Mean relative difference: 1.345755" [4] "Component 12588:
> Mean relative difference: 0.5620998" [5] "Component 12589: Mean
> relative difference: 0.6626076" [6] "Component 12590: Mean relative
> difference: 0.487681" [7] "Component 12591: Mean relative
> difference: 0.4553995" [8] "Component 12592: Mean relative
> difference: 0.4263176" [9] "Component 12593: Mean relative
> difference: 0.5712674" [10] "Component 12594: Mean relative
> difference: 0.9872221" [11] "Component 12595: Mean relative
> difference: 0.6511257" [12] "Component 12596: Mean relative
> difference: 1.118144" [13] "Component 12597: Mean relative
> difference: 0.7570665" [14] "Component 12598: Mean relative
> difference: 0.6555009" [15] "Component 12599: Mean relative
> difference: 0.8429751" [16] "Component 12600: Mean relative
> difference: 0.8220698" [17] "Component 12601: Mean relative
> difference: 0.5806326" [18] "Component 12602: Mean relative
> difference: 0.9459446" [19] "Component 12603: Mean relative
> difference: 0.005017876"
> [20] "Component 12604: Mean relative difference: 0.002521591"
> [21] "Component 12605: Mean relative difference: 17.80169" [22]
> "Component 12606: Mean relative difference: 0.6895796" [23]
> "Component 12607: Mean relative difference: 1.435084" [24]
> "Component 12608: Mean relative difference: 0.9511139" [25]
> "Component 12609: Mean relative difference: 0.004859382"
> [26] "Component 12610: Mean relative difference: 0.002441555"
> [27] "Component 12611: Mean relative difference: 0.004895062"
> [28] "Component 12612: Mean relative difference: 0.02192048"
> [29] "Component 12613: Mean relative difference: 0.00498039"
> [30] "Component 12614: Mean relative difference: 0.002502659"
> [31] "Component 12615: Mean relative difference: 0.009985646"
> [32] "Component 12616: Mean relative difference: 0.004943459"
> [33] "Component 12617: Mean relative difference: 0.002484009"
> [34] "Component 12618: Mean relative difference: 0.009911417"
> [35] "Component 12619: Mean relative difference: 0.004907072"
> [36] "Component 12620: Mean relative difference: 0.002465635"
> [37] "Component 12621: Mean relative difference: 0.01475742"
> [38] "Component 12622: Mean relative difference: 0.002423802"
> [39] "Component 12623: Mean relative difference: 0.01259747"
> [40] "Component 12624: Mean relative difference: 0.01005289"
> [41] "Component 12625: Numeric: lengths (20, 69) differ"
>
> indexMaster[12586]
> $`AFFX-hum_alu_at`
> [1] 250676 375731 211930 327074 352651 352650 43467 159539 159540 171051
> [11] 231270 102022 263133 140413 278581 178848 178847 214557 191494 191495
> [21] 38016 228673 277142 275902 114699 215783 77671 21489 21490 267349
> [31] 276339 257960 305449 305450 279710 22853 22854 163552 336192 339896
> [41] 35483 43466 43465 171053 171052 102024 102023 263134 278582 278583
> [51] 277394 373143 214556 58516 58515 270948 249046 318231 126406 57395
> [61] 247739 154374 154375 154376 144365 254115 254116 337311 337310
>
> indexSlave[[1]][12586]
> $`AFFX-HUMGAPDH/M33197_3_at`
> [1] 121056 36917 261006 35492 126329 176294 141533 368370 155742 84276
> [11] 137751 66002 4753 26735 71484 102044 214559 25066 127497 114997
>
> sessionInfo()
> R version 2.6.0 (2007-10-03)
> x86_64-unknown-linux-gnu
> locale:
> LC_CTYPE=de_DE.UTF-8;LC_NUMERIC=C;LC_TIME=de_DE.UTF-8;LC_COLLATE=de_DE.UTF-8;LC_MONETARY=de_DE.UTF-8;LC_MESSAGES=de_DE.UTF-8;LC_PAPER=de_DE.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=de_DE.UTF-8;LC_IDENTIFICATION=C
> attached base packages:
> [1] tools stats graphics grDevices utils datasets
> methods [8] base other attached packages:
> [1] hgu95av2cdf_2.0.0 affydata_1.11.3 affy_1.16.0 [4]
> preprocessCore_1.0.0 affyio_1.6.1 Biobase_1.16.1 [7]
> Rmpi_0.5-6 snow_0.2-9 loaded via a namespace (and
> not attached):
> [1] rcompgen_0.1-15
>
>
> clusterEvalQ(c1, sessionInfo())
> [[1]]
> R version 2.6.0 (2007-10-03)
> x86_64-unknown-linux-gnu
> locale:
> C
> attached base packages:
> [1] tools stats graphics grDevices utils datasets
> methods [8] base other attached packages:
> [1] hgu95av2cdf_2.0.0 affy_1.16.0 preprocessCore_1.0.0
> [4] affyio_1.6.1 Biobase_1.16.1 snow_0.2-9 [7]
> Rmpi_0.5-6
>
>
> stopCluster(c1)
>
>
>
> --
> Dipl.-Tech. Math. Markus Schmidberger
>
> Ludwig-Maximilians-Universität München
> IBE - Institut für medizinische Informationsverarbeitung,
> Biometrie und Epidemiologie
> Marchioninistr. 15, D-81377 Muenchen
> URL: http://ibe.web.med.uni-muenchen.de Mail: Markus.Schmidberger [at]
> ibe.med.uni-muenchen.de
> Tel: +49 (089) 7095 - 4599
>
--
Martin Morgan
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109
Location: Arnold Building M2 B169
Phone: (206) 667-2793
More information about the Bioconductor
mailing list