Re: [BioGeoBEARS] Error in tree

578 views
Skip to first unread message
Message has been deleted

Nick Matzke

unread,
Feb 15, 2017, 5:50:11 PM2/15/17
to bioge...@googlegroups.com
Re-sending to group since I think my reply didn't go through...

On Thu, Feb 16, 2017 at 11:41 AM, Nick Matzke <nick....@anu.edu.au> wrote:
Hi Uzma,

Your files load without error and seem to physically run in BioGeoBEARS, however, that phylogeny does not look like a time-tree.  It ought to be a time-tree!  See: http://phylo.wikidot.com/biogeobears-mistakes-to-avoid#use_dated_phylogenies

It's also possible that having all of those tiny short branches (molecular branchlengths I assume) is causing numerical problems leading to inconsistent crashes.

Also make sure you source() the patched code after **every time you use the library(BioGeoBEARS) command**. See: http://phylo.wikidot.com/biogeobears-mistakes-to-avoid#source_everything2

Cheers, Nick



On Thu, Feb 16, 2017 at 1:28 AM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
I have made the tree file and geography file but I don't I got error every time, sometimes in geography file, sometime in tree, could you please check my files?

Thank you,
Best,
Uzma 

--
You received this message because you are subscribed to the Google Groups "BioGeoBEARS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.
To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/7c514290-a44e-48d4-9c9f-63bb6685fb08%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


uzm...@fjwu.edu.pk

unread,
Feb 16, 2017, 8:17:50 AM2/16/17
to BioGeoBEARS
Thank you Nick, I want to do Biogeographical Stochastic Mapping (BSM) on my data for ancestral area reconstruction. I need to know what kind of phylogeny tree is needed like it should be  dated and what other files are required besides geography data. Is there any example script that I could follow and understand. 
Best,
Uzma

Nick Matzke

unread,
Feb 16, 2017, 2:56:58 PM2/16/17
to bioge...@googlegroups.com
On Fri, Feb 17, 2017 at 2:17 AM, <uzm...@fjwu.edu.pk> wrote:
Thank you Nick, I want to do Biogeographical Stochastic Mapping (BSM) on my data for ancestral area reconstruction. I need to know what kind of phylogeny tree is needed like it should be  dated and what other files are required besides geography data. Is there any example script that I could follow and understand. 
Best,
Uzma


Yes. Please study the BioGeoBEARS help pages on Phylowiki, they will answer all basic questions:

Basic analysis:

BSM:

Cheers, Nick


 


On Wednesday, February 15, 2017 at 5:28:37 PM UTC+5, uzm...@fjwu.edu.pk wrote:
Hi Nick,
I have made the tree file and geography file but I don't I got error every time, sometimes in geography file, sometime in tree, could you please check my files?

Thank you,
Best,
Uzma 

--
You received this message because you are subscribed to the Google Groups "BioGeoBEARS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.
To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 9, 2017, 8:13:38 PM3/9/17
to BioGeoBEARS
Hello Nick, I'm running BSM and every time I got this error:
Error in matrix(data = NA, nrow = ntips, ncol = nareas) :    invalid 'ncol' value (too large or NA) In addition: Warning message: In read_PHYLIP_data(lgdata_fn) : NAs introduced by coercion. 
I don't know why it is so, please help me out in this. 
Best, 
Uzma 


On Friday, February 17, 2017 at 12:56:58 AM UTC+5, Nick Matzke wrote:
On Fri, Feb 17, 2017 at 2:17 AM, <uzm...@fjwu.edu.pk> wrote:
Thank you Nick, I want to do Biogeographical Stochastic Mapping (BSM) on my data for ancestral area reconstruction. I need to know what kind of phylogeny tree is needed like it should be  dated and what other files are required besides geography data. Is there any example script that I could follow and understand. 
Best,
Uzma


Yes. Please study the BioGeoBEARS help pages on Phylowiki, they will answer all basic questions:

Basic analysis:

BSM:

Cheers, Nick


 


On Wednesday, February 15, 2017 at 5:28:37 PM UTC+5, uzm...@fjwu.edu.pk wrote:
Hi Nick,
I have made the tree file and geography file but I don't I got error every time, sometimes in geography file, sometime in tree, could you please check my files?

Thank you,
Best,
Uzma 

--
You received this message because you are subscribed to the Google Groups "BioGeoBEARS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears...@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

Nick Matzke

unread,
Mar 9, 2017, 8:22:24 PM3/9/17
to bioge...@googlegroups.com
On Fri, Mar 10, 2017 at 12:13 PM, <uzm...@fjwu.edu.pk> wrote:
Hello Nick, I'm running BSM and every time I got this error:
Error in matrix(data = NA, nrow = ntips, ncol = nareas) :    invalid 'ncol' value (too large or NA) In addition: Warning message: In read_PHYLIP_data(lgdata_fn) : NAs introduced by coercion. 
I don't know why it is so, please help me out in this. 
Best, 
Uzma 


Hi Uzma,

Please read the error message: "nareas" means number of areas.  So, the "ncol" input for the matrix() command is getting "NA" from "nareas".

The number of areas is counted from the geography file (lgdata_fn, FYI it "lgdata_fn" means "lagrange-formatted data file name").

But, you can see the warning: "Warning message: In read_PHYLIP_data(lgdata_fn) : NAs introduced by coercion."

This suggests some kind of problem with your geography data file.  The most likely issue is that you aren't actually referencing the correct file location (the path, and the filename).  So, make sure that you are correctly referring to the file location..

If that is fixed, then make sure that the file is properly formatted (e.g. no extra blanks at the end of the line, every species has the correct number of numbers in the 2nd column, etc.).

Short version: run the command "read_PHYLIP_data(lgdata_fn)" by itself, and get that to work properly, before proceeding.

Cheers, 

 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 12, 2017, 2:52:49 PM3/12/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
I have checked the location and file format both are fine but don't know why I got error every time. 
Error in matrix(data = NA, nrow = ntips, ncol = nareas) : 
  invalid 'ncol' value (too large or NA)
In addition: Warning message:
In read_PHYLIP_data(lgdata_fn) : NAs introduced by coercion

Best,
Uzma

Nick Matzke

unread,
Mar 13, 2017, 8:24:06 PM3/13/17
to bioge...@googlegroups.com
Hi Uzma,

Please send the file and I can take a look.  Cheers, Nick


To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 14, 2017, 3:49:48 PM3/14/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Thank you, here you go. I also did DEC analysis using Lagrange, is there any way we could past the results on tree as you know we get text file as output in Lagrange? 
Best,
Uzma
ca_concatenated_beast_2017_03_08.NEWICK
ecoregion_beast_2017_03_08.DATA

Nick Matzke

unread,
Mar 14, 2017, 7:24:16 PM3/14/17
to bioge...@googlegroups.com
On Wed, Mar 15, 2017 at 6:49 AM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
Thank you, here you go. I also did DEC analysis using Lagrange, is there any way we could past the results on tree as you know we get text file as output in Lagrange? 
Best,
Uzma

It's actually waaaaay simpler to just run the (identical*) DEC model in BioGeoBEARS than it is to hack something together to get the Lagrange outputs (which one? Python or Lagrange) onto a nice phylogeny graphic in R.  I did it, but it would take me awhile to figure it out again. I posted everything I did here, though, if you want to try figuring it out again.

Processing output from DIVA, Python Lagrange, and C++, and graphical comparison to BioGeoBEARS

To have any hope of understanding how the different programs output their results, you have to understand the different node numbering schemes in the different programs:

Comparing node numbers from APE/BioGeoBEARS, Python Lagrange, C++ Lagrange, and DIVA

Cheers, Nick

PS: Details...

(*) The likelihood calculation and ML parameter inferences are identical** between Lagrange DEC and BioGeoBEARS DEC, but the ancestral state probabilities can differ a bit, because:

Although under the DEC model, the ML log-likelihood and parameter estimates are identical, the ancestral state probabilities are not exactly the same because:

=========================
Lagrange estimated node state/range probabilities under "local optimization", i.e. for every possible ancestral range at every node, Lagrange runs a new - ML optimization of the parameters while fixing that range at that node.
- BioGeoBEARS estimates ancestral states/ranges under the globally optimum ML model.
=========================


(**) There seem to be some differences in time-stratified analyses, for the C++ Lagrange, but I've never gotten to the bottom of it. See Matzke 2014, Systematic Biology (BioGeoBEARS LnL for a stratified analysis agreed with Python Lagrange-stratified but not C++ Lagrange-stratified).

(***) Neither Python nor C++ Lagrange will necessarily treat branch lengths appropriately for non-ultrametric trees (tree where not all of the tips come up to the present, e.g. trees with fossils. BioGeoBEARS handles these correctly (although it's always possible I've screwed something up, I've worked pretty hard on this bit, including the hardest case, a time-stratified tree with fossils in it). 



 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

Nick Matzke

unread,
Mar 14, 2017, 7:43:33 PM3/14/17
to bioge...@googlegroups.com
On Wed, Mar 15, 2017 at 6:49 AM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
Thank you, here you go. I also did DEC analysis using Lagrange, is there any way we could past the results on tree as you know we get text file as output in Lagrange? 
Best,
Uzma

On Tuesday, March 14, 2017 at 5:24:06 AM UTC+5, Nick Matzke wrote:
Hi Uzma,

Please send the file and I can take a look.  Cheers, Nick

Hi! Thanks for the file, that helps A TON in diagnosing issues.

I found the problem with this geography file...

The file started with:

=======================================
183 4(P N I H)
Carex_acuta 1000
Carex_acutiformis 1001
[...]
=======================================

...the error message produced was (for google-ability):
=======================================
geogfn="/drives/GDrive/z_help/Uzma_help/rebiogeobearsreerrorintree/ecoregion_beast_2017_03_08.DATA"
tipranges = getranges_from_LagrangePHYLIP(lgdata_fn= geogfn)

Read 5 items
Error in matrix(data = NA, nrow = ntips, ncol = nareas) : 
  invalid 'ncol' value (too large or NA)
In addition: Warning message:
In read_PHYLIP_data(lgdata_fn) : NAs introduced by coercion
=======================================

I fixed the file by adding a tab before the "(P N I H)" (as shown in the example file!!)
=======================================
183 4 (P N I H)
Carex_acuta 1000
Carex_acutiformis 1001
[...]
=======================================

...then it works fine...
=======================================
geogfn="/drives/GDrive/z_help/Uzma_help/rebiogeobearsreerrorintree/ecoregion_beast_2017_03_08.DATA"
tipranges = getranges_from_LagrangePHYLIP(lgdata_fn= geogfn)

tipranges = getranges_from_LagrangePHYLIP(lgdata_fn= geogfn)
Read 6 items
Read 366 items

tipranges
An object of class "tipranges" (from the global environment)
numeric(0)
Slot "df":
                        P N I H
Carex_acuta             1 0 0 0
Carex_acutiformis       1 0 0 1
[...]
=======================================

Cheers!
Nick












 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 16, 2017, 9:24:28 PM3/16/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Thank you for your kind reply. I used Lagrange python script. I parsed the files according to link you shared but got two states probability although I have 4 ranges. I'm attaching results here please see and suggest me if these are fine. 
Best,
Uzma
LGpy_plots_carex_concatenated_2017_03_15.pdf

Nick Matzke

unread,
Mar 16, 2017, 11:13:58 PM3/16/17
to bioge...@googlegroups.com
On Fri, Mar 17, 2017 at 12:24 PM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
Thank you for your kind reply. I used Lagrange python script. I parsed the files according to link you shared but got two states probability although I have 4 ranges. I'm attaching results here please see and suggest me if these are fine. 
Best,
Uzma


Hi Uzma,

As the title of last plot says, the pie charts are giving the probability of the most probable state, and the other pie slice is the summed probability of everything else. So it's not identical to a BioGeoBEARS plot (and the colors might not mean the same thing between different nodes).

If I recall correctly, more information is saved inside the R output.  But also IIRC, Lagrange doesn't even report the probability of all possible states, just the top 95% of the probability, so getting a perfect table of everything isn't possible. 

Cheers, Nick




 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 17, 2017, 11:23:00 AM3/17/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Thank you Nick. So, if I want to do DEC analysis using BioGeoBEARS then how could I do, I mean is there any script on website please share the link?
Thank you,
Best,
Uzma

Nick Matzke

unread,
Mar 18, 2017, 12:23:08 AM3/18/17
to bioge...@googlegroups.com
On Sat, Mar 18, 2017 at 2:22 AM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
Thank you Nick. So, if I want to do DEC analysis using BioGeoBEARS then how could I do, I mean is there any script on website please share the link?
Thank you,
Best,
Uzma


Hi Uzma,

DEC is the very first model in the example script!  I'm sure you've already run it!! 

Nick

 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 18, 2017, 8:51:13 AM3/18/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Yes, but if we create the input file using Lagrange configurator and then is there any possibility that we could run that file in R using BioGeoBEARS, so instead of parsing Lagrange results using BioGeoBEARS if we could use input file of Lagrange and then run in R using BioGeoBEARS.
Thank you,
Best,
Uzma  

Nick Matzke

unread,
Mar 19, 2017, 2:07:04 AM3/19/17
to bioge...@googlegroups.com
On Sat, Mar 18, 2017 at 11:51 PM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick,
Yes, but if we create the input file using Lagrange configurator and then is there any possibility that we could run that file in R using BioGeoBEARS, so instead of parsing Lagrange results using BioGeoBEARS if we could use input file of Lagrange and then run in R using BioGeoBEARS.
Thank you,
Best,
Uzma

Hi -- 

No, there is no way to run BioGeoBEARS with the inputs from Python Lagrange.  You will have to manually construct input text files that contain the same/similar information as what you gave Python Lagrange.

The BioGeoBEARS input files do (mostly) overlap with the input files for C++ Lagrange.

Example input files are all here:

Cheers,
Nick


 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 21, 2017, 10:34:41 AM3/21/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick, I have run DEC and DEC+J analyses using BioGeoBEARS, and it ran nicely, In the last step I got this message:
> system(cmdstr) # Plot it
Warning messages:
1: In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA,  :
  unused control arguments ignored
2: In optimx.check(par, optcfg$ufn, optcfg$ugr, optcfg$uhess, lower,  :
  Parameters or bounds appear to have different scalings.
  This can cause poor performance in optimization. 
  It is important for derivative free methods like BOBYQA, UOBYQA, NEWUOA.
3: In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA,  :
  unused control arguments ignored
what does it mean?
Thanks,
Best,
Uzma

Nick Matzke

unread,
Mar 21, 2017, 7:30:44 PM3/21/17
to bioge...@googlegroups.com
Hi -- you can basically ignore those warnings:

Although, you will want to check if your maximum likelihood optimizations make sense, since the optimization can sometimes fail to optimize:

ML optimization routines and their pitfalls

But if it makes sense, you should be good.

Cheers!
Nick


To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/728cbaa4-0ec8-4055-9eb2-5ffda942825f%40googlegroups.com.

uzm...@fjwu.edu.pk

unread,
Mar 22, 2017, 2:16:34 PM3/22/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Thank you! so if I want to do BSM from results of these two analyses (DEC and DEC+J), how could I do, if I want to use output of these two analyses for BSM analyses. Because in phylo.wikidot website there is a complete script start from DEC analysis then DEC+J and then ancestral states and then stochastic mapping. I have done upto ancestral states and just want to do BSM analysis, how could I do this any script just for it.
Thank you,
Best,
Uzma

uzm...@fjwu.edu.pk

unread,
Mar 28, 2017, 12:55:53 PM3/28/17
to BioGeoBEARS, mat...@nimbios.org
Hi all,
I'm plotting per area probability and didn't understand this command in script:
# DEC
titletxt = "Psychotria: probability of occupancy per area\nunder DEC M0 analysis"

what does this file name mean?
when I run the script I got this error:

Error in file(filename, "r", encoding = encoding) : 
  cannot open the connection

Thank you,
Uzma

Nick Matzke

unread,
Mar 28, 2017, 4:29:32 PM3/28/17
to bioge...@googlegroups.com


On Wednesday, March 29, 2017, <uzm...@fjwu.edu.pk> wrote:
Hi all,
I'm plotting per area probability and didn't understand this command in script:
# DEC
titletxt = "Psychotria: probability of occupancy per area\nunder DEC M0 analysis"

what does this file name mean?


It's not a file name. It's the text of a title for a plot. Thus, "titletxt".


 
when I run the script I got this error:

Error in file(filename, "r", encoding = encoding) : 
  cannot open the connection

Thank you,
Uzma

This means that whatever filename is being referred to in your script is not being found. 

This has come up several times- the fundamental skill you need here is to figure out working directories (wd) by use of:

 getwd
 setwd
list.files

...And connecting that in your head to looking at files and directories with your computer's file manager (Eg Finder on a Mac). 

Everything will seem confusing until you figure this out, and probably every time you get a "cannot open the connection" error this is the issue.

Finally, to get better help online, you should ALWAYS paste the exact line of the script that caused the error, then the error. Otherwise we are just guessing about what script and what command you are using.

Cheers, nick 



 
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears+unsubscribe@googlegroups.com.

To post to this group, send email to bioge...@googlegroups.com.
Visit this group at https://groups.google.com/group/biogeobears.

uzm...@fjwu.edu.pk

unread,
Mar 28, 2017, 6:26:34 PM3/28/17
to BioGeoBEARS, mat...@nimbios.org
Hi, I got this error although I always set my directory
Error in file(filename, "r", encoding = encoding) : 
  cannot open the connection
In addition: Warning messages:
1: In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA,  :
  unused control arguments ignored
2: In optimx.check(par, optcfg$ufn, optcfg$ugr, optcfg$uhess, lower,  :
  Parameters or bounds appear to have different scalings.
  This can cause poor performance in optimization. 
  It is important for derivative free methods like BOBYQA, UOBYQA, NEWUOA.
3: In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA,  :
  unused control arguments ignored
4: In file(filename, "r", encoding = encoding) :
  cannot open file 'C:\Dropbox\_njm\__packages\BioGeoBEARS_setup\inst\extdata\a_scripts\plot_phylo3_nodecoords.R': No such file or directory
Thank you,
Uzma

Nick Matzke

unread,
Mar 28, 2017, 8:22:32 PM3/28/17
to bioge...@googlegroups.com

To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/1157a3d8-7f01-4166-a06b-c8f64102aa26%40googlegroups.com.

uzm...@fjwu.edu.pk

unread,
Mar 29, 2017, 11:06:06 AM3/29/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick, thank you. I have a question about "time stratified analysis", what does it mean? and how we could make time stratified table usually require in BSM analysis and I also got this error as master time stratified table is not present. I'm sorry for a very introductory question.
Best,
Uzma

uzm...@fjwu.edu.pk

unread,
Mar 29, 2017, 4:05:22 PM3/29/17
to BioGeoBEARS, mat...@nimbios.org
Hi, got this error, when I added timeperiod:

Error in if (nodes_to_chainsaw[i] <= numtips) { : 
  missing value where TRUE/FALSE needed

Nick Matzke

unread,
Mar 29, 2017, 8:04:45 PM3/29/17
to bioge...@googlegroups.com
On Thu, Mar 30, 2017 at 2:06 AM, <uzm...@fjwu.edu.pk> wrote:
Hi Nick, thank you. I have a question about "time stratified analysis", what does it mean? and how we could make time stratified table usually require in BSM analysis and I also got this error as master time stratified table is not present. I'm sorry for a very introductory question.
Best,
Uzma

Hi Uzma,

The purpose of software help groups is not to give everyone a personal one-on-one tutorial on introductory material. I will give you an answer here, but I expect that in the future you will put in serious work to understand the methods and the software.  I know it is difficult, and with free biology software put out by researchers the documentation is often far from perfect. But, you will get further faster by doing the background reading and then working to puzzle out (by googling, reading articles, etc.) than by asking intro questions.

That said, here is the basic answer to your question: 

- Time-stratification just means that you have a series of time-bins, with different geography in each time bin. 
- The different geography is specified by a dispersal-multiplier matrix, a distance matrix, etc.
- A times file is also required
- The formats for these files are here: http://phylo.wikidot.com/biogeobears#texteditors
- Search the example script on "stratified" to see the bits you would change to run a time-stratified analysis.
- After you've successfully run a time-stratified likelihood analysis, you can then use that as an input to a time-stratified BSM analysis
- If you want to run BSM on a non-stratified likelihood analysis, then you should set stratified=FALSE in the BSM script (if I recall correctly).


But, overall, some general comments:

1. Before proceeding further with biogeography analyses, I STRONGLY RECOMMEND that you READ, CAREFULLY AND SLOWLY, Ree & Smith 2008 (Systematic Biology), Matzke 2013 (Frontiers in Biogeography), and Matzke 2014 (Systematic Biology). If things like time-stratification are mysterious, then you probably are not yet in a position to understand what you are doing in any biogeography analysis, and any research you produce will suffer from that.

Time-stratification was introduced first in Ree & Smith (2008), and is explained there.



2. Then, work through the example scripts SLOWLY AND CAREFULLY.  In other words:

WRONG IDEA: just copy-and-paste and then post questions when it doesn't work
RIGHT IDEA: copy-and-paste line-by-line, READING EACH COMMENT CAREFULLY AND THINKING ABOUT IT so you understand what is going on


3. Then, read these links, to clarify some of the issues that many people get confused, often because they come from a parsimony background and don't have much background in probabilistic methods:



4. Ways to get better help:

Weak question: Please explain Y to me from scratch, I haven't googled it or word-searched the help pages or anything

Strong question: Can someone give me a reference, or link, that explains Y?  I have googled it, and also looked in Google Scholar, so I understand up to point X, but item Y is still puzzling.

Cheers,
Nick



 

uzm...@fjwu.edu.pk

unread,
Mar 29, 2017, 8:24:44 PM3/29/17
to BioGeoBEARS, mat...@nimbios.org
Hi Nick,
Thank you so much for a very detailed answer. Yes I think I'm not in a position to work on time stratified but if BSM work with non-stratified then it would be fine for me, I tried non-stratified analysis and I'll try again after reading all these stuff as I got error every time in running that. Yes  I have read Rick Ree papers when I have done LAGRANGE analysis.
Thanks again!
Best,
Uzma 
Reply all
Reply to author
Forward
0 new messages