beta_diversity_through_plots.py "result contains negative eigenvalues."

[XiaoYu Liu]

I just updated QIIME to 1.9.0. When I run beta diversity plot script:

beta_diversity_through_plots.py -f -i otus/otus_sorted.biom -m map_key.txt -t otus/rep_set.tre -e 8450 -p qiime_params.txt -a -O 8 -o bdiv

I got following error, and folder bdiv/bray_curtis_emperor_pcoa_plot is empty:

/mnt/software/anaconda/envs/qiime1.9/lib/python2.7/site-packages/skbio/stats/ordination/_principal_coordinate_analysis.py:107: RuntimeWarning: The result contains negative eigenvalues. Please compare their magnitude with the magnitude of some of the largest positive eigenvalues. If the negative ones are smaller, it's probably safe to ignore them, but if they are large in magnitude, the results won't be useful. See the Notes section for more details. The smallest eigenvalue is -0.0149050441596 and the largest is 0.892552539716.

Why did this happen, and how can I get around it?

Thanks!

[Daniel McDonald]

Hi Xiao,

You can find more information here:

http://scikit-bio.org/docs/latest/generated/generated/skbio.stats.ordination.PCoA.html#skbio.stats.ordination.PCoA

It is likely safe to ignore this as it looks like they're small in magnitude. This situation existed in prior versions of qiime but we were not flagging it for users.

Best,

Daniel

--

---
You received this message because you are subscribed to the Google Groups "Qiime Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qiime-forum...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Clayton Collings]

I also got the same error, but when I run Qiime-1.8.0, I do not.

/group/bioinfo/apps/apps/Qiime-1.9.0/python/lib/python2.7/site-packages/skbio/stats/ordination/_principal_coordinate_analysis.py:107: RuntimeWarning: The result contains negative eigenvalues. Please compare their magnitude with the magnitude of some of the largest positive eigenvalues. If the negative ones are smaller, it's probably safe to ignore them, but if they are large in magnitude, the results won't be useful. See the Notes section for more details. The smallest eigenvalue is -0.00402124290445 and the largest is 0.221597918062.

  RuntimeWarning

[habr...@ucdavis.edu]

Hello,

I am using Qiime 1.9 and I also get the "Run Time" Warning. Initially I tried:

beta_diversity_through_plots.py -i otu_table_NO_singletons.biom -o beta_diversity_even17224/ -t rep_set.tre -m MappingFile_Compost_corrected.txt -e 17224

17224 is the minimum sequence value from the biom_summary.txt but since it failed I tried repeating the command with a lower -e value and I tried not using the -e element at all. Both attempts failed. Is this a result of the computer on which I am working or am I using the command wrong?

Thank you.

-Hilary

[xzron...@gmail.com]

 Hi,I am using Qiime 1.9 and I have the same error.If you have  solved the problem, can you tell me how to do? Thank you!

在 2015年2月25日星期三 UTC+8下午12:30:26，QIIME noobie写道：

[Greg Caporaso]

This is not an error, but rather a warning. See the warning message that is printed - it describes the issue, and how to determine if your results are negatively affected by it. 

 

[Jose]

Dear Greg and other QIIME developers,

If it is only a warning why the script does not go on? Please find attached my log with the description of the problem (QIIME 1.9 was used but this is also happening in 1.8 although the log is not quite the same). In my case I am using core_diversity_analyses.py and I simply cannot finalize the script. First time this happens to me, I look forward to hearing more about this issue (what causes it, how I can avoid it, etc). The other users were talking about beta_diversity_through_plots.py, I assume related scripts will also be affected by this??

Thanks!

Jose

[Colin Brislawn]

Hello Jose, 

Did you read your log file, looking for clues? I found two important clues near the end. 

I can see that principal_coordinates.py finished (with this warning about eigenvalues). Afterwards, the script make_emperor.py failed (with this error "cp: cannot create regular file `...': Operation not permitted".

While you also received a similar warning about negative eigenvalues, your script crashed for a different reason. Could you start a new thread for your "cp: cannot create regular file" error? Thanks!

Colin

[Jose]

Thanks Colin! I will start a new the thread for it.

Jose

[nathaniel hubert]

I had the same problem - any suggestions?

I did get PCoA, alpha and beta diveristy analyses that look complete for the most part, but the script never completed.

MacQIIME nathaniels-MacBook-Pro:karen_041116 $ core_diversity_analyses.py -i /Users/nathanielhubert/Desktop/karen_041116/Lewis_SD_otu_table_mc2_w_tax_no_pynast_failures.biom -o core_diversity_Lewis_SD_otu_table_mc2_w_tax_no_pynast_failures/ -m /Users/nathanielhubert/Desktop/karen_041116/041216_Karen_Lewis_SD_merge_MAP.txt -e 14942 -t /Users/nathanielhubert/Desktop/karen_041116/open_ref_otus_Lewis_SD/rep_set.tre -c newTRT_Trial -aO 3

/macqiime/anaconda/lib/python2.7/site-packages/skbio/stats/ordination/_principal_coordinate_analysis.py:107: RuntimeWarning: The result contains negative eigenvalues. Please compare their magnitude with the magnitude of some of the largest positive eigenvalues. If the negative ones are smaller, it's probably safe to ignore them, but if they are large in magnitude, the results won't be useful. See the Notes section for more details. The smallest eigenvalue is -0.0159448047287 and the largest is 1.11046710756.

  RuntimeWarning

/macqiime/anaconda/lib/python2.7/site-packages/matplotlib/tight_layout.py:225: UserWarning: tight_layout : falling back to Agg renderer

  warnings.warn("tight_layout : falling back to Agg renderer")

[nathaniel hubert]

I'm sorry, it did finish eventually, looks like everything is normal. What is a large magnitude for negative eigen vectors?