Hi Keith, really nice videos. But I have a problem. I have all my data as you , also grouped eg 3 groups . Wenn I select a column and then click on the stack coloured rows into columns does not happent what happens to you. I do not take 3 columns (my samples) for that one column (e.g the x Parasite) . Instead all my data is lost and what i get is that the selecte column stays only and the salmpels remain as they are. Is it a bug? I also have a question. what is the difference between "stack al colourd raws into columns ) and grouped columns to multivar or row to multivar? thank you alot. Keep up the great job
@kamcgnt
6 жыл бұрын
First thing may be a bug. But I cannot really diagnose software problems at a distance from a few sentences. Sorry. For the second question, I don't know: read the manual and experiment.
@paleouy
10 жыл бұрын
Dear Keith: Thanks a lot for your answer! I have a doubt concerning the use of var-covar vs. correlation matrix in PCA. My variables are a group of indexes determined from real measures. Some of them fluctuate between 0 and 1. And others acquire more wide values. Therefore, although all of them are continuous variables, obviously they are of different kinds. So my question is if in this case would be more adequate the use of the correlation matrix instead the var-covar matrix. Or if I need some special procedure of normalization. On the other hand. Theoretically speaking, do you think it is methodologically correct combine continuous and discrete variables in a same data matrix? If this is possible, how can you do that? Normalizing them? I have a pool of some accessory, discrete variables, but I don’t know if they could be useful to me… Thanks a lot in advance! And thanks a lot for share these very useful videos! Greetings form Uruguay.
@paleouy
10 жыл бұрын
Hi Keith! Could you tell me in which of yours videos could I learn how to create groups of data? Thanks in advance!
@kamcgnt
10 жыл бұрын
Any video with ANOVA, PERMANOVA or ANOSIM, as they all require groups. PAST Version 2 does groups in different ways for different analyses. Haven't checked this in Version 3.
@yayamal1
8 жыл бұрын
Dear Keith How to check the data are significant in PAST. i do that by pearson correlation in spss
@kamcgnt
8 жыл бұрын
+Hatem Hassan You can do correlations in PAST: on the Stats menu.
@yadavghimirey
8 жыл бұрын
Hi Keith! I am using PAST 3. I have my data of benthic organisms taken from 12 plots.. The 12 plots are located in such a way that the first four plots are located at a distance of 50 m, second four plots at 100 m and the third four plots at 150 m from sea shore. So there are basically three zones (Groups) based on the increasing distance from sea shore. I have created a column zones that has Z1, Z2 and Z3 with their corresponding abundance data for species. I have also given the colours to the three different zones. Now I want to do Anosim and Simper tests but I am still confused on how to classify three groups because PAST tells me that there should be atleast two or more groups when I try to run these tests.
@kamcgnt
8 жыл бұрын
+Panthera Pardus Hi! See most recent video. Your problems are due (I think) to a change between PAST 2 and 3 in how groups are defined.
@yadavghimirey
8 жыл бұрын
That's great! Will try to do that.. The software though seems to be good :) Thank you very much!
Пікірлер: 10