# Pairwise-differences - metric
**URL:** https://forum.qiime2.org/t/pairwise-differences-metric/6836
**Category:** User Support
**Created:** [November 7, 2018, 3:48pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836 "2018-11-07T15:48:28Z")
**Posts on this page:** 6
**Page:** 1
### Author:  [@Imindias](https://forum.qiime2.org/u/Imindias)
#### Post date: [November 7, 2018, 3:48pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/1 "2018-11-07T15:48:28Z")
Continuing the discussion from [--p-metric for beta diversity](https://forum.qiime2.org/t/p-metric-for-beta-diversity/4415/8):
Hi everybody,
I need more clarifications about this topic. I can't understand why is possible to perform this analysis with shannon and faith\_pd metrics, but for example with evenness and observed\_otus, not.
I have tried to perform the metadata tabulate and it is ok. I'm figuring out that the problem is the word that we have to put in --p-metric, because although the --m-metadata-file works, I think that the problem is in --p-metric.
Could anybody help me?
thanks a lot,
Isabel
---
### Author:  [@Nicholas\_Bokulich](https://forum.qiime2.org/u/Nicholas_Bokulich)
#### Post date: [November 7, 2018, 3:51pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/2 "2018-11-07T15:51:05Z")
> [@Imindias](#):
>
> I need more clarifications about this topic. I can’t understand why is possible to perform this analysis with shannon and faith\_pd metrics, but for example with evenness and observed\_otus, not.
shannon and faith\_pd have transparent metric names.
evenness is labeled "pielou\_e" or something along those lines in the alpha diversity artifact — using `metadata tabulate` on the alpha diversity artifact should clarify the correct name.
If you are still having trouble, please post the exact command you are attempting to use, as well as QZVs (output of `metadata tabulate`) for all files you are using as input.
Thanks!
---
### Author:  [@Imindias](https://forum.qiime2.org/u/Imindias)
#### Post date: [November 7, 2018, 4:23pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/3 "2018-11-07T16:23:27Z")
Great! I`ve got it with alpha diversity indexes, but and with the case of PCoA? which is the word that we should use? I've tried pcoa, unifrac, unweighted\_unifrac, but...
This is my command:
qiime longitudinal pairwise-differences --m-metadata-file dieta\_vs\_bs\_map\_var.txt --m-metadata-file /Volumes/Elements/dieta\_vs\_BSFinal/bien-con-guion/core-metrics-results-even22272/unweighted\_unifrac\_pcoa\_results.qza --p-metric pcoa --p-group-column Treatment --p-state-column Time --p-state-1 pre --p-state-2 post --p-individual-id-column Individual --p-replicate-handling random --o-visualization dietbs-pairwise-differences-unweighted\_pcoa\_results.qzv
and this is the warning:
**Plugin error from longitudinal:**
**metric must be a valid metadata or feature table column.**
Thanks a lot,
Isabel
---
### Author:  [@Nicholas\_Bokulich](https://forum.qiime2.org/u/Nicholas_Bokulich)
#### Post date: [November 7, 2018, 5:38pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/4 "2018-11-07T17:38:32Z")
Use `qiime metadata tabulate` on this file: `/Volumes/Elements/dieta_vs_BSFinal/bien-con-guion/core-metrics-results-even22272/unweighted_unifrac_pcoa_results.qza` to see what the column names are. I believe it should be something like "Axis 1".
---
### Author:  [@Imindias](https://forum.qiime2.org/u/Imindias)
#### Post date: [November 7, 2018, 8:12pm UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/5 "2018-11-07T20:12:50Z")
Great! it works!
thanks a lot!!
---
### Author:  [@system](https://forum.qiime2.org/u/system)
#### Post date: [December 9, 2018, 2:12am UTC](https://forum.qiime2.org/t/pairwise-differences-metric/6836/6 "2018-12-09T02:12:52Z")
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.