[R-sig-eco] Unbalanced data and random effects
Thank you very much for these explanations. It is quite technical and I am not sure that I got it all, but I will try to find the book of GelmanHill to get more insight into shrinkage. I read the book of Zuur and as you said the topic is not extensively covered. ___ Les prévisions météo pour aujourd'hui, demain et jusqu'à 8 jours ! Voila.fr http://meteo.voila.fr/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] Unbalanced data and random effects
Dear all, I performed a census of insects at different sites and measured there size. I would like to know if size is related to an environmental factor. I modelled the size as a fonction of the factor with site as a random variable to account for within-site variability. However I have strong unbalanced data with some sites having only two individuals and others up to 100. Is having site as a random factor sufficient to deal with this strong data unbalance? The residual fit of the data is quite bad, certainly because of the strong difference in variance among sites. Would anybody have some advice? Thank you! ___ Les prévisions météo pour aujourd'hui, demain et jusqu'à 8 jours ! Voila.fr http://meteo.voila.fr/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] null model for testing nestedness
Thank you very much. Yes it is working with oecosimu, exept that it does not seem to work for weighted data. There is the possibility to specify weighted = TRUE: oecosimu(matrix,nestednodf, method = quasiswap, nsimul = 999, order = FALSE, weighted =TRUE) However, I get only null values and p=1. For weighted = F, I get good values. Best wishes ___ Quiz TV : Vous êtes fan de la série Friends ? 5 questions ici http://tv.voila.fr/quiz/quiz-special-friends_14538959.html ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] null model for testing nestedness
Dear Jari, Thank you very much for this clear answer. I did not get that quasiswap only concerned binary data. After reading your explanations, I think I'll stay to binary data and avoid the issue of weighted ones, which are much less straightforward to interpret. Anyway, I will have a look at the development versions. Best wishes, Valérie Message du 25/09/13 à 15h45 De : Jari Oksanen A : Copie à : Objet : Re: [R-sig-eco] null model for testing nestedness Valerie, There are at least two problems here: the way you call oecosimu() and how nestdnodf(..., weighted =TRUE) works with binary data. If you specify a *binary* null model as method, then you will get binary data. Even if you supplied quantitative data, they are transformed into 1/0 (presence/absence) data. You specified method = quasiswap, and that is binary model. Another problem is that nestednodf(..., weighted = TRUE) seems to evaluate the statistics all as zeros if you request weighted (= quantitative data) analysis of non-quantitative data (binary). It cannot perform weighted analysis if there are no weights, but still I think it should return something else than zeros. We'll have a look at that issue. You should specify a non-binary null model if you want to have a non-binary (weighted) analysis. Quantitative null models are problematic, and vegan release version does not have much choice here. I think r2dtable may be the only one. Development version of vegan in http://www.r-forge.r-project.org/ has a wider gamme of non-binary null models, but I think you need to be brave to use quantitative null models. They are something for people who are not afraid of going to areas where angels fear to tread. FWIW, weighted nestednodf seems to work in oecosimu if you ask for a quantitative nullmodel (r2dtable in my tests) both with the release version (2.0-8 or 2.0-9) and with the development version (2.1-35 or 2.1-36). But you really need to to specify a quantitative null model. Both null models and oecosimu are completely re-written and re-designed in development versions. Cheers, Jari Oksanen On 25/09/2013, at 15:56 PM, wrote: Thank you very much. Yes it is working with oecosimu, exept that it does not seem to work for weighted data. There is the possibility to specify weighted = TRUE: oecosimu(matrix,nestednodf, method = quasiswap, nsimul = 999, order = FALSE, weighted =TRUE) However, I get only null values and p=1. For weighted = F, I get good values. Best wishes ___ Quiz TV : Vous êtes fan de la série Friends ? 5 questions ici http://tv.voila.fr/quiz/quiz-special-friends_14538959.html ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology ___ Quiz TV : Vous êtes fan de la série Friends ? 5 questions ici http://tv.voila.fr/quiz/quiz-special-friends_14538959.html ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] null model for testing nestedness
Dear all, I would like to implement a null model to test if nestedness of a matrix departs from chance. There is an example in package bipartite with the function nullmodel: obs - unlist(networklevel(web, index=weighted NODF)) nulls - nullmodel(web, N=100, method=1) null - unlist(sapply(nulls , networklevel, index=weighted NODF)) This works well, however, I have the impress that prior to apply the function networklevel, the initial matrix is being reordered to achieve maximal packing. However, I don't want my matrix to be reordered, but I did not manage to find how to specify it. In the package vegan, there is the function nestednodf with the option order=FALSE, but I could not implement a null model based on this function (because the output contain multiple attributes). Any help welcomed ___ Quiz TV : Vous êtes fan de la série Friends ? 5 questions ici http://tv.voila.fr/quiz/quiz-special-friends_14538959.html ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] dispersion parameter in binomial model
Dear all, I computed a binomial model with a proportion as response variable using glm(cbind(realized, not realized)~x,family=binomial). The output tells me that the dispersion parameter taken is 1. For comparison I computed the same model using family=quasibinomial and I get a dispersion parameter of 0.5. The resultats are very different between the two models and in regard to the plotted data, the quasibinomial model seems to be more accurate. I am a bit confused about how to know if my data are accurately fitted by a binomial model or if they are under- or overdispersed and I'd rather use the binomial or another fitting model. I found this formula to calculate the dispersion parameter, but I am not sure if it is accurate for a binomial model: phi=sum(((realized/(realized+not realized))-model$fitted)^2/model$fitted)/model$df.residual With this formula I get for both the binomial and the quasibinomial model a phi=0.4. Is this a sign of underdispersion? Thank you! ___ Qu'y a-t-il ce soir à la télé ? D'un coup d'œil, visualisez le programme sur Voila.fr http://tv.voila.fr/programmes/chaines-tnt/ce-soir.html ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] adonis and temporal changes
Dear all, I would like to test changes in species dissimilarity matrices over time, taking into account that the measurement are repeated in each site over years. I used the adonis function: adonis(diss.matrix~year, strata=site). However if I do the same model entering site as an additional fixed effect (this was applied this way in a paperI read): adonis(diss.matrix~site+year, strata=site), I get exactly the same estimate for year, but the variance explained is much higher. I am a bit lost regarding how much of the variance in dissimilarity is really explained by temporal changes. Thank you very much ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx?product=StEdCH2K2Af=-3000 ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] (no subject)
Thank you for these explanations. If I put strata=site, this means that for each site my dissimilarity matrix of year 1 and year 2 will be permuted and the observed changes compared to these random permutation? Adding site as a fixed factor then ensure that I am testing changes in time site by site. Am I correct? To my design: I have 30 permanent sites, 10 of each category of isolation (Isolation = factor with 3 levels: 3x10 sites = 30 sites). I conducted the samples in three years in each site. I have thus 1 sampling (species composition) pro site pro year. I would like to know how the sampled communities change with time, either on a site basis, or at the level of isolation (I may compare multi-site dissimilarity among isolation levels between years). I am not really interested in knowing what proportion of differences in species community is due to space vs time, but I would like to really focus on the temporal changes. That's why I think putting site as a fixed effect should be appropriate. But if you have any suggestion or think this is not correct, I would be pleased to have your opinion. Cheers, Valerie On 18/02/2013, at 14:04 PM, Pierre THIRIET wrote: Dear Valérie, If I remember well, your design includes: Isolation categories: 3 levels Sites: nested within Isolation categories (10 levels, a total of 30 sites) How many replicates per site and time? Time:? how many years you have? Only one sampling per year? Within sites and years, samples were random or it is always exactly the same area you sample (e.g. permanent quadrats)? for adonis, consider that strata is for constraining permutations, which is different than terms in the formulae. Exactly. The 'strata' only influence the permutations and have no effect in formula nor effect defined in the formula. Currently the 'strata' are the only way to constrain the permutations. However, in the R-Forge version of vegan and in vegan 2.2-0 (to be released in April) you can give a permutation matrix as an input to adonis. You can generate the permutation matrix with, say, shuffleSet function of the permute package. This allows generation of restricted permutations for instance for time series. Vegan command vegandocs(permutations) will open up the vignette of the permute package for your inspection, and this will give some examples of defining restricted permutations. At some timeframe we are completely moving to the permute package, but you can already use its permutation matrices as input with these new and upcoming versions of vegan from R-Forge. Cheers, Jari -- Jari Oksanen, Dept Biology, Univ Oulu, 90014 Finland ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx?product=StEdCH2K2Af=-3000 ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] adonis and temporal changes
Thank you for these explanations. If I put strata=site, this means that for each site my dissimilarity matrix of year 1 and year 2 will be permuted and the observed changes compared to these random permutation? Adding site as a fixed factor then ensure that I am testing changes in time site by site. Am I correct? To my design: I have 30 permanent sites, 10 of each category of isolation (Isolation = factor with 3 levels: 3x10 sites = 30 sites). I conducted the samples in three years in each site. I have thus 1 sampling (species composition) pro site pro year. I would like to know how the sampled communities change with time, either on a site basis, or at the level of isolation (I may compare multi-site dissimilarity among isolation levels between years). I am not really interested in knowing what proportion of differences in species community is due to space vs time, but I would like to really focus on the temporal changes. That's why I think putting site as a fixed effect should be appropriate. But if you have any suggestion or think this is not correct, I would be pleased to have your opinion. Cheers, Valerie On 18/02/2013, at 14:04 PM, Pierre THIRIET wrote: Dear Valérie, If I remember well, your design includes: Isolation categories: 3 levels Sites: nested within Isolation categories (10 levels, a total of 30 sites) How many replicates per site and time? Time:? how many years you have? Only one sampling per year? Within sites and years, samples were random or it is always exactly the same area you sample (e.g. permanent quadrats)? for adonis, consider that strata is for constraining permutations, which is different than terms in the formulae. Exactly. The 'strata' only influence the permutations and have no effect in formula nor effect defined in the formula. Currently the 'strata' are the only way to constrain the permutations. However, in the R-Forge version of vegan and in vegan 2.2-0 (to be released in April) you can give a permutation matrix as an input to adonis. You can generate the permutation matrix with, say, shuffleSet function of the permute package. This allows generation of restricted permutations for instance for time series. Vegan command vegandocs(permutations) will open up the vignette of the permute package for your inspection, and this will give some examples of defining restricted permutations. At some timeframe we are completely moving to the permute package, but you can already use its permutation matrices as input with these new and upcoming versions of vegan from R-Forge. Cheers, Jari -- Jari Oksanen, Dept Biology, Univ Oulu, 90014 Finland ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx? product=StEdCH2K2Af=-3000 ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx?product=StEdCH2K2Af=-3000 ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Re: [R-sig-eco] adonis and temporal changes
Dear Steve, Thank you very much. I do not exactly understand why the test for isolation will be wrong, would you have some some explanation? In a linear regression, you cannot assess the effect of single variable if the interaction (in which your variable is part) is significant. So if I get a significant result for the isolation*year effect I should conclude that there is an interaction between isolation and year. If the interaction is not significant, should I drop it to get the correct estimate for the year effect? I would have an additional question: I have also an environemental gradient (continuous, one value pro site, constant over the years). Is it possible to include it? Best wishes Valerie Message du 18/02/13 à 15h41 De : Steve Brewer A : v_coudr...@voila.fr, r-sig-ecology@r-project.org Copie à : Objet : Re: [R-sig-eco] adonis and temporal changes Valerie, Adonis does not define fixed or random effects, and you therefore cannot define multiple error terms. However, if your model statement looks something like this - isolation*year + site, strata = site - then you will get the correct test for the isolation x year interaction and the correct test for the year effect. The test for isolation will be wrong, because the residual error is used to test all effects, when it is only appropriate for testing the year effect and the year * isolation interaction. The isolation between-subjects effect should be tested with the site effect but is not. The key point is here to make strata = site and to NOT specify the site- interactions with isolation or year. In this way, site will be treated as a block for the within-subjects effects and thus could be considered a random effect. Hope this helps. J. Stephen Brewer Professor Department of Biology PO Box 1848 University of Mississippi University, Mississippi 38677-1848 Brewer web page - http://home.olemiss.edu/~jbrewer/ FAX - 662-915-5144 Phone - 662-915-1077 On 2/18/13 8:19 AM, v_coudr...@voila.fr wrote: Thank you for these explanations. If I put strata=site, this means that for each site my dissimilarity matrix of year 1 and year 2 will be permuted and the observed changes compared to these random permutation? Adding site as a fixed factor then ensure that I am testing changes in time site by site. Am I correct? To my design: I have 30 permanent sites, 10 of each category of isolation (Isolation = factor with 3 levels: 3x10 sites = 30 sites). I conducted the samples in three years in each site. I have thus 1 sampling (species composition) pro site pro year. I would like to know how the sampled communities change with time, either on a site basis, or at the level of isolation (I may compare multi-site dissimilarity among isolation levels between years). I am not really interested in knowing what proportion of differences in species community is due to space vs time, but I would like to really focus on the temporal changes. That's why I think putting site as a fixed effect should be appropriate. But if you have any suggestion or think this is not correct, I would be pleased to have your opinion. Cheers, Valerie On 18/02/2013, at 14:04 PM, Pierre THIRIET wrote: Dear Valérie, If I remember well, your design includes: Isolation categories: 3 levels Sites: nested within Isolation categories (10 levels, a total of 30 sites) How many replicates per site and time? Time:? how many years you have? Only one sampling per year? Within sites and years, samples were random or it is always exactly the same area you sample (e.g. permanent quadrats)? for adonis, consider that strata is for constraining permutations, which is different than terms in the formulae. Exactly. The 'strata' only influence the permutations and have no effect in formula nor effect defined in the formula. Currently the 'strata' are the only way to constrain the permutations. However, in the R-Forge version of vegan and in vegan 2.2-0 (to be released in April) you can give a permutation matrix as an input to adonis. You can generate the permutation matrix with, say, shuffleSet function of the permute package. This allows generation of restricted permutations for instance for time series. Vegan command vegandocs(permutations) will open up the vignette of the permute package for your inspection, and this will give some examples of defining restricted permutations. At some timeframe we are completely moving to the permute package, but you can already use its permutation matrices as input with these new and upcoming versions of vegan from R-Forge. Cheers, Jari -- Jari Oksanen, Dept Biology, Univ Oulu, 90014 Finland ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx? product=StEdCH2K2Af=-3000
Re: [R-sig-eco] adonis and temporal changes
Thank you very much for your explanations. I still would have a question about the proportion explained. I got for example an R2 of 0.11 for years*isolation and 0.46 for site. Does this mean that most of the variation in species composition is between sites and within site variation (also from one year to the other) is relatively small? And what about the 40% unexplained...I do not well see where variation can be if it is neither between nor within sites. Many thanks Best, Valerie Message du 18/02/13 à 21h49 De : Steve Brewer A : v_coudr...@voila.fr, r-sig-ecology@r-project.org Copie à : Objet : Re: [R-sig-eco] adonis and temporal changes Valerie, If I understand your design correctly, you're doing a repeated measures analysis, in which isolation is a between-subjects (I.e., between-sites) effect. Year and the year x isolation interaction are within-subjects effects. Because repeated measurements on composition are being taken on the same site in three years, you use strata to restrict the permutation within each site as if site were were a random block containing the different years of measurement. Accordingly, there should be two error terms: site(isolation) to test the isolation main effect, and the site*year(isolation), which in this case is equivalent to the residual error, which is the appropriate error term for testing the year effect and the year x isolation interaction. The test for isolation is wrong because adonis cannot use more than one error term to test effect and thus is using the residual error to test all effects. It should use the site(isolation) term to test the isolation effect, but it does not. Using the residual error to test the isolation effect amounts to pseudoreplication. It assumes that the three measurements of composition in different years on the same site are independent observations. They are not. Often, however, people are not interested in the between-subjects effects (in this case, the main effect of isolation). Rather they are interested in the interaction with time (in this case, isolation x year). I don't see that you are justified in pooling any term with the error term just because it is not significant. Again, the problem is pseudoreplication. You're treating correlated observations as if they were independent observations. Pooling the isolation x year interaction with the residual error term artificially inflates your error df even more. I'm afraid I don't know R well enough to explain how to analyze the covariate. J. Stephen Brewer Professor Department of Biology PO Box 1848 University of Mississippi University, Mississippi 38677-1848 Brewer web page - http://home.olemiss.edu/~jbrewer/ FAX - 662-915-5144 Phone - 662-915-1077 On 2/18/13 1:49 PM, v_coudr...@voila.fr wrote: Dear Steve, Thank you very much. I do not exactly understand why the test for isolation will be wrong, would you have some some explanation? In a linear regression, you cannot assess the effect of single variable if the interaction (in which your variable is part) is significant. So if I get a significant result for the isolation*year effect I should conclude that there is an interaction between isolation and year. If the interaction is not significant, should I drop it to get the correct estimate for the year effect? I would have an additional question: I have also an environemental gradient (continuous, one value pro site, constant over the years). Is it possible to include it? Best wishes Valerie Message du 18/02/13 à 15h41 De : Steve Brewer A : v_coudr...@voila.fr, r-sig-ecology@r-project.org Copie à : Objet : Re: [R-sig-eco] adonis and temporal changes Valerie, Adonis does not define fixed or random effects, and you therefore cannot define multiple error terms. However, if your model statement looks something like this - isolation*year + site, strata = site - then you will get the correct test for the isolation x year interaction and the correct test for the year effect. The test for isolation will be wrong, because the residual error is used to test all effects, when it is only appropriate for testing the year effect and the year * isolation interaction. The isolation between-subjects effect should be tested with the site effect but is not. The key point is here to make strata = site and to NOT specify the site- interactions with isolation or year. In this way, site will be treated as a block for the within-subjects effects and thus could be considered a random effect. Hope this helps. J. Stephen Brewer Professor Department of Biology PO Box 1848 University of Mississippi University, Mississippi 38677-1848 Brewer web page - http://home.olemiss.edu/~jbrewer/ FAX - 662-915-5144 Phone - 662-915-1077 On 2/18/13 8:19 AM, v_coudr...@voila.fr wrote: Thank you for these explanations. If I put
[R-sig-eco] binomial regression with non integers
Dear Pierre, Thank you very much for your answer. In fact I would like to make two different analyses: one spatial and one temporal. For the spatial analysis, I will compute the dissimilarities in the way you suggested it, using beta.pair and dbRDA. For temporal analysis of beta diversity between sites, Baselga proposed to use the function beta.temp that produces for each site a value for beta1, beta2 and betaTotal. I have 30 sites, 10 of factor level 1, 10 of factor level 2 and 10 of factor level 3. I thought that the best way to look at the relationships between the factor and the components of beta diversity was to make a logistic regression as mentioned earlier: glm(cbind(beta1,beta2)~x,family=quasibinomial). However since beta1 and beta2 are non-integers I am not sure about being allowed to use binomial regression. I would like to mention as well that using family=binomial I get a warning about non-integer values, whereas by using family=quasibinomial no such warning appears. My model being not overdispersed, there would be no justification of using a quasi-model. But maybe somebody may have some more information about this. Thank you Valérie ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx?product=StEdCH2K2Af=-3000 ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] binomial regression with non integers
Dear all, I am investigating diversity in different sites. I partitioned my measure of diversity into to additive components (Baselga 2012) and get for each site a value of overal diversity change (between 0 and 1) and a value for each additive component, such that for each site beta1+beta2=beta_total. I would like to make a regression model to test if the proportion of diversity due to beta1 (beta1/beta_total) is signifcantly different according to an explanatory factor. If beta1 had been an integer value, I would have used a binomial model. However, since beta1 is not an integer I don't think that I am allowed to use the formel glm(cbind(beta1,beta2)~x,family=binomial)? What alternative method could I use? I hope that my question is not too confuse. Thank you very much. Valérie ___ C'est l'année du Serpent ! Connaissez-vous votre signe astral chinois ? Découvrez-le ici http://astrocenter.voila.fr/voila/Presentation.aspx?product=StEdCH2K2Af=-3000 ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Re: [R-sig-eco] proportion data with many zeros
Thank you very much for clarifying this point. My algorithm is certainly pretty bad because as you say I am basically looking at zeros. One point I don't really understand is that for a pollen type I have a lot of pollen collected at date 1, some at time 2, few at time 3 and not at all at time 4. I get a significant difference between time 1 and 2 but no significance between 1 and 3 or 1 and 4. That is illogical...maybe is it anyway a problem of the residuals because the residuals are pretty well balanced for time points with fitted values 0, but for time points with no pollen collected there is no variance at all. Well I think that if I had a very large number of data such that the non-zero part of my data would look nicely continuous I could use some zero-inflated models, but with only 4 points in time and a positive part of the model which does not fit well a continuous distribution it is difficult. I'd certainly better take a descriptive way of presenting my data for sparse pollen types. Best wishes Valérie Message du 04/02/13 à 13h15 De : Liz Pryde A : v_coudr...@voila.fr Copie à : Objet : Re: [R-sig-eco] proportion data with many zeros Hi, If you're using a categorical predictor those QQ plots Etc are pretty useless. Just do a residuals vs fits plots and make sure the residuals look Randomly scattered. Is the problem with the smaller pollen types just that they're very low across all time scales? The algorithm won't fit b/c you're basically looking at zero data - or a vector of zeroes. So you can assume that this is sig diff from the abundant types. This is to do with the way ML estimation works - it's a bit complicated. Some people suggest using bayes methods for this ( it works well) but its way too over-complicated for what you're trying to answer. The mean variance relationship is specified by the 'family' part if the GLM formula. It is essentially the error structure if your data. Liz On 04/02/2013, at 7:55 PM, v_coudr...@voila.fr wrote: I tried to use tweedie and it again worked very well for the most abundant pollen types and when trying to fit the less abundant ones I got the error: glm.fit: algorithm did not converge. I have the impress that it is hopeless to try fitting a model...But anyway thank you very much for making me aware of tweedie. I still should go a bit more into the theorical background. I just wonder about the residuals. For the pollen types that can be modelled, the QQ-plots don't look very nice, but the residuals are relatively well homogeneously distributed. It is difficult to judge how good the fit is, but the results make sense in regard to the raw data. Valérie ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] adonis and negative F-values
Dear all, I used adonis to perform a test of the pairwise site dissimilarity indices proposed by Baselga (2010, 2012) in the package betapart. I am concerned about my results because I get some negative F-values. I read in another post that this may happen because of the presence of negative eigenvalues. However I was wondering if this does invalidate the results, or if they are still interpretable in some way. Moreover in case the results are still valid, do you think that providing a result table containing negative F-values will be considered for publication or be an argument of refusal? I may use distance-based RDA with the cailliez correction instead, would it be a good alternative to adonis for testing the effect of a three-level factor on the dissimilarity measures? Best wishes, Valerie Coudrain ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Re: [R-sig-eco] proportion data with many zeros
Thank you Liz, I don't know tweedie, I'll have a look at it, but I have indeed some high values. I know about the problems linked to the arcsine transformation. I won't consider it anyway. I'd like to use either the raw values of pollen grain counts or a logistic quasibinomial model. Best, Valérie Message du 02/02/13 à 20h47 De : Liz Pryde A : v_coudr...@voila.fr Copie à : Cade Brian , r-sig-ecology@r-project.org Objet : Re: [R-sig-eco] proportion data with many zeros Have you plotted the raw data to have a look at the distribution? You could try another exponential family distribution like tweedie that has a mass at zero but is otherwise similar to poisson/gamma - so you're directly modeling the zeroes. It won't work if you have a lot of high values though. Proportions are tricky. Have a read of the Warton paper (2012/11?) the arcsine is asinine. Liz On 02/02/2013, at 6:34 PM, v_coudr...@voila.fr wrote: Thank you very much for this suggestion. In fact I reconsidered my question and I am not sure that zero-inflated model is what I need. If I understood it properly, a zero-inflated model is best suited when we don't know if zero values are true or false absences (right?). In my case all zero values are assumed to be real absence and are therefore informative. However, fitting quasipoisson on raw counts or quasibinomial on proportion gives me awful distributions of residuals and meaningless results. Valérie Message du 01/02/13 à 17h22 De : Cade, Brian A : v_coudr...@voila.fr Copie à : r-sig-ecology@r-project.org Objet : Re: [R-sig-eco] proportion data with many zeros For a fully parametric approach, you might want to use of zero-inflated beta distribution (e.g., as available in gamlss package), which is designed for zero-inflated proportions. Or for a semi-parametric approach, you could estimated a sequence of quantile regression estimates (e.g., in package quantreg), where some interval (hopefully not to large) of the quantiles will be uninformative because they are massed at the zero values. Brian Brian S. Cade, PhD U. S. Geological Survey Fort Collins Science Center 2150 Centre Ave., Bldg. C Fort Collins, CO 80526-8818 email: brian_c...@usgs.gov tel: 970 226-9326 On Fri, Feb 1, 2013 at 1:30 AM, wrote: Dear all, I am trying to test how the proportion of pollen of different plants found in the brood cells of a wild bee changes over time. I conducted 4 sampling sessions (thus time is a factor with 4 levels) and collected several pollen samples for each time point (300 pollen grains counted for each sample). I thought about applying a quasi-binomial glm: y = cbind(total pollen - pollen of plant X, pollen of plant X) glm(y~time, family=quasibinomial) The problem is that I have a lot of zero value, because the pollen of some plants only occurred rarely or very clumped in time. I thought about applying a zero-inflated model, but I have never used it and I am not sure if it is suitable for proportion data. Additionally I wondered if I have to consider the fact that I don't have the same number of pollen sample for each date, which makes my design unbalanced. Thank you in advance for advice. Best wishes Valérie ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[R-sig-eco] proportion data with many zeros
Dear all, I am trying to test how the proportion of pollen of different plants found in the brood cells of a wild bee changes over time. I conducted 4 sampling sessions (thus time is a factor with 4 levels) and collected several pollen samples for each time point (300 pollen grains counted for each sample). I thought about applying a quasi-binomial glm: y = cbind(total pollen - pollen of plant X, pollen of plant X) glm(y~time, family=quasibinomial) The problem is that I have a lot of zero value, because the pollen of some plants only occurred rarely or very clumped in time. I thought about applying a zero-inflated model, but I have never used it and I am not sure if it is suitable for proportion data. Additionally I wondered if I have to consider the fact that I don't have the same number of pollen sample for each date, which makes my design unbalanced. Thank you in advance for advice. Best wishes Valérie ___ CAN 2013 : résultats et matchs en direct à suivre sur Voila.fr http://sports.voila.fr/football/can/ ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Re: [R-sig-eco] Difference between mantel test and adonis
Thank you very much for your explanations. So if I understood it correctly, a significant outcome from adonis() (or capscale()) should indicate me that my species community changes along the gradient? Nice that you also mention betadisper, because I'd like also to look at variation in species composition within the factor levels and along the gradient and came about the same issue. Best wishes Valérie Message du 16/01/13 à 10h08 De : syro...@sci.muni.cz A : v_coudr...@voila.fr Copie à : r-sig-ecology@r-project.org Objet : Re: [R-sig-eco] Difference between mantel test and adonis Hi Valérie, adonis is analogous to RDA or CCA, as it directly estimates the variance in the distance matrix attributable to an independent variables(s), with the advantage that one may use any distance measure. It parallels r2 in a linear model. Mantel test simply calculates the correlation between two distance matrices (and tests it via permutations), thus, one gets the idea about wthether there is a linear relationship between them at all. Cheers, Vit Dear Martin, Tank you very much. I thought about constrained ordination, but the distance matrix I am using is not among the usual (Euclidian, Bray-Curtis,...). An option would be to use distance-based RDA, which is almost the same as adonis, but I read that adonis should be even better (?) Anyway I am mainly interested to understand the difference between analyses based on distance matrix of the environemental gradient (Mantel test), or on the gradient directly. Best, Valérie Message du 16/01/13 à 00h53 De : Martin Weiser A : v_coudr...@voila.fr Copie à : r-sig-ecology@r-project.org Objet : Re: [R-sig-eco] Difference between mantel test and adonis v_coudr...@voila.fr píše v Út 15. 01. 2013 v 16:08 +0100: Dear sig-eco users, I would like to investigate the changes in a species community along an ecological gradient. I first thought about performing a Mantel test and infer if differences in species composition are related to differences in the ecological gradient. I noticed that the function adonis (package vegan) could handle continuous variables as well. However, the ecological gradient is not entered as a distance matrix and therefore I don't understand exactly how to interpret the outcome of the adonis test and what is the difference to the Mantel test. Any help will be apprectiate. Best wishes. Valerie ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology Dear Valerie, I would go for constrained ordination: it is easily interpretable and tailored exactly to your needs. To do so, in R use ade4 or vegan (I am not familiar with ade4) and run CCA or RDA (depends on data and taste). Plus, Jari Oksanen wrote easy-to-understand tutorial: http://cc.oulu.fi/~jarioksa/opetus/metodi/vegantutor.pdf Another great free resources are Mike Palmer's website: http://ordination.okstate.edu/ and ordnews mailinglist: ordn...@colostate.edu I hope this helps. Best, Martin W. ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Re: [R-sig-eco] Difference between mantel test and adonis
Dear Martin, Tank you very much. I thought about constrained ordination, but the distance matrix I am using is not among the usual (Euclidian, Bray-Curtis,...). An option would be to use distance-based RDA, which is almost the same as adonis, but I read that adonis should be even better (?) Anyway I am mainly interested to understand the difference between analyses based on distance matrix of the environemental gradient (Mantel test), or on the gradient directly. Best, Valérie Message du 16/01/13 à 00h53 De : Martin Weiser A : v_coudr...@voila.fr Copie à : r-sig-ecology@r-project.org Objet : Re: [R-sig-eco] Difference between mantel test and adonis v_coudr...@voila.fr píše v Út 15. 01. 2013 v 16:08 +0100: Dear sig-eco users, I would like to investigate the changes in a species community along an ecological gradient. I first thought about performing a Mantel test and infer if differences in species composition are related to differences in the ecological gradient. I noticed that the function adonis (package vegan) could handle continuous variables as well. However, the ecological gradient is not entered as a distance matrix and therefore I don't understand exactly how to interpret the outcome of the adonis test and what is the difference to the Mantel test. Any help will be apprectiate. Best wishes. Valerie ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology Dear Valerie, I would go for constrained ordination: it is easily interpretable and tailored exactly to your needs. To do so, in R use ade4 or vegan (I am not familiar with ade4) and run CCA or RDA (depends on data and taste). Plus, Jari Oksanen wrote easy-to-understand tutorial: http://cc.oulu.fi/~jarioksa/opetus/metodi/vegantutor.pdf Another great free resources are Mike Palmer's website: http://ordination.okstate.edu/ and ordnews mailinglist: ordn...@colostate.edu I hope this helps. Best, Martin W. ___ Les téléviseurs écrans plats sont en soldes sur le comparateur de prix de Voila.fr http://shopping.voila.fr/vitrine/televiseurs-ecran-plat ___ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
