and now...Mac version for OS/X!

[cooch]

The latest and greatest version of MARK (7.1) has been compiled for Mac (OS/X) users. Since most Mac types will use the Mac-compiled version with RMark, the details for installing OS/X MARK (a.k.a. MacMARK) for use with RMark on a Mac are posted here:

http://www.phidot.org/software/mark/rmark/linux/

Please read the details carefully. The version of the MARK executable (32-bit only for the moment) was compiled to be as 'platform neutral' as possible, but that is not always easy to fully accomplish. In particular, if you're running a machine with a non-Intel chip (is that still possible for Macs?), and have problems, (i) don't be surprised, and (ii) let us know.

Reminder that MacMARK is 'used by you, and supported by you'. If it works for you -- great, awesome, verging on totally rad. If it doesn't, you're on your own.

Further, if you can't get it to play nice with RMark, then post a query in the RMark subforum (i.e., its all Jeff's fault )

[jlaake]

Thanks Evan. That was a nice present for all of the Mac users. Linkages between RMark and outside executables such as mark.exe and file viewer are the possible problems that may be encountered. I don't own a Mac so it will be hard for me to debug problems but I'll ask some of my colleagues to try it and we'll see what happens. There is no OsX binary on CRAN so you'll have to build the package from source.

regards --jeff

[cooch]

Thanks Evan. That was a nice present for all of the Mac users.

Well, I recently inherited a ~2-3 year old Mac PowerBook, so I figured, what the heck. Since Gary ported MARK to gfortran, I figured it was doable. Turns out, was very straightforward. Only major limit is that my Mac is 32-bit. But, now that I've got things set up, all I need to do is find someone I know with a newer, spiffier (presumably 64-bit machine), and I'll compile for that at some point. But, the 32-bit version works just fine (at least, from the command line....).

Linkages between RMark and outside executables such as mark.exe and file viewer are the possible problems that may be encountered. I don't own a Mac so it will be hard for me to debug problems but I'll ask some of my colleagues to try it and we'll see what happens.

No doubt. I have a Mac, but wouldn't even know where to start with such issues (I'm still trying to figure out how to get rid of the jumping icons on the bottom of the screen. The late great Steve Jobs might have liked it, but I want all the 'cute hipster crap' off my desktop.

There is no OsX binary on CRAN so you'll have to build the package from source.

Presumably this will be relatively straightforward for most users.

If I get bored over the holiday, I might give it a try.

[cooch]

Following might be of some help in getting folks up and running with RMark on a Mac:

viewtopic.php?f=21&t=1050&p=2862&hilit=mac#p2862

[bacollier]

Well, I recently inherited a ~2-3 year old Mac PowerBook, so I figured, what the heck. Since Gary ported MARK to gfortran, I figured it was doable. Turns out, was very straightforward. Only major limit is that my Mac is 32-bit. But, now that I've got things set up, all I need to do is find someone I know with a newer, spiffier (presumably 64-bit machine), and I'll compile for that at some point. But, the 32-bit version works just fine (at least, from the command line....).

Which I assume is your (futile, late, and inexpensive) effort to increase your cool and trendy quotient...

Linkages between RMark and outside executables such as mark.exe and file viewer are the possible problems that may be encountered. I don't own a Mac so it will be hard for me to debug problems but I'll ask some of my colleagues to try it and we'll see what happens.

No doubt. I have a Mac, but wouldn't even no where to start with such issues (I'm still trying to figure out how to get rid of the jumping icons on the bottom of the screen. The late Steve Jobs might have like it, but I want all the 'cute hipster crap' off my desktop.

and

Presumably this will be relatively straightforward for most users.

If I get bored over the holiday, I might give it a try.

I just tried a build on my wife's MacBook and saw massive crashes left and right with with RMark build routine (best I can tell) and what looks to be administrative access to the mark.exe location. I did not see any MARK install issues until her computer puked on the screen, pinwheeled for about 5 minutes, and then effectively went and hid. My best bug report is "oh &^%$" at this point... I will try it on a clean MacBook tomorrow that does not have any admin limitations at the office and see if I can recreate the issue and track it down any better.

FWIW, I just set up the Linux 64 bit on a updated Ubuntu and Mint, no issues thus far.

\bret

[cooch]

Which I assume is your (futile, late, and inexpensive) effort to increase your cool and trendy quotient...

Man, I'm waaay past the level of 'futile' for that objective.

I just tried a build on my wife's MacBook and saw massive crashes left and right with with RMark build routine (best I can tell) and what looks to be administrative access to the mark.exe location. I did not see any MARK install issues until her computer puked on the screen, pinwheeled for about 5 minutes, and then effectively went and hid. My best bug report is "oh &^%$" at this point... I will try it on a clean MacBook tomorrow that does not have any admin limitations at the office and see if I can recreate the issue and track it down any better.

Probably because you didn't download R or MacMARK through iTunes.

However, one step at a time -- did the executable work at least? You can try if from a terminal. Take the following code -- call it mcmcdipper.inp. The, try running MARK from the command line:

mark i=mcmcdipper.inp o=mcmcdipper.lst

Code: Select all

proc title test2;
proc chmatrix occasions=7 groups=2 etype=Live NoHist hist=300;
   glabel(1)=Group 1;
   glabel(2)=Group 2;
   time interval 1 1 1 1 1 1;
/* European Dipper Data, Live Recaptures, 7 occasions, 2 groups
 Group 1=Males Group 2=Females */
1111110  1 0 ;
1111100  0 1 ;
1111000  1 0 ;
1111000  0 1 ;
1101110  0 1 ;
1100000  1 0 ;
1100000  1 0 ;
1100000  1 0 ;
1100000  1 0 ;
1100000  0 1 ;
1100000  0 1 ;
1010000  1 0 ;
1010000  0 1 ;
1000000  1 0 ;
1000000  1 0 ;
1000000  1 0 ;
1000000  1 0 ;
1000000  1 0 ;
1000000  0 1 ;
1000000  0 1 ;
1000000  0 1 ;
1000000  0 1 ;
0111111  0 1 ;
0111111  0 1 ;
0111110  0 1 ;
0111100  1 0 ;
0111100  0 1 ;
0111100  0 1 ;
0111000  1 0 ;
0111000  0 1 ;
0110110  0 1 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  1 0 ;
0110000  0 1 ;
0110000  0 1 ;
0110000  0 1 ;
0110000  0 1 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  1 0 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0100000  0 1 ;
0011111  0 1 ;
0011111  0 1 ;
0011110  1 0 ;
0011110  0 1 ;
0011100  1 0 ;
0011100  1 0 ;
0011100  1 0 ;
0011100  1 0 ;
0011100  0 1 ;
0011100  0 1 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  1 0 ;
0011000  0 1 ;
0011000  0 1 ;
0011000  0 1 ;
0011000  0 1 ;
0010110  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  1 0 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0010000  0 1 ;
0001111  1 0 ;
0001111  1 0 ;
0001111  1 0 ;
0001111  1 0 ;
0001111  1 0 ;
0001111  1 0 ;
0001111  0 1 ;
0001111  0 1 ;
0001110  1 0 ;
0001110  1 0 ;
0001110  1 0 ;
0001110  0 1 ;
0001110  0 1 ;
0001110  0 1 ;
0001110  0 1 ;
0001100  1 0 ;
0001100  1 0 ;
0001100  1 0 ;
0001100  1 0 ;
0001100  1 0 ;
0001100  1 0 ;
0001100  0 1 ;
0001100  0 1 ;
0001100  0 1 ;
0001100  0 1 ;
0001100  0 1 ;
0001011  0 1 ;
0001001  1 0 ;
0001001  0 1 ;
0001000  1 0 ;
0001000  1 0 ;
0001000  1 0 ;
0001000  1 0 ;
0001000  1 0 ;
0001000  1 0 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0001000  0 1 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  1 0 ;
0000111  0 1 ;
0000111  0 1 ;
0000111  0 1 ;
0000111  0 1 ;
0000111  0 1 ;
0000111  0 1 ;
0000110  1 0 ;
0000110  1 0 ;
0000110  1 0 ;
0000110  0 1 ;
0000110  0 1 ;
0000110  0 1 ;
0000110  0 1 ;
0000110  0 1 ;
0000110  0 1 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  1 0 ;
0000100  0 1 ;
0000100  0 1 ;
0000100  0 1 ;
0000100  0 1 ;
0000100  0 1 ;
0000100  0 1 ;
0000100  0 1 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  1 0 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000011  0 1 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  1 0 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000010  0 1 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  1 0 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
0000001  0 1 ;
proc estimate MCMC MCMCCSVFile=MCMC.CSV link=Logit varest=2ndPart ;
     model={Phi(g*t) p(g*t) PIM 22 RMARK 24};
     group=1 Phi rows=6 cols=6 Triang Time=1;
     group=2 Phi rows=6 cols=6 Triang Time=7;
     group=1 p rows=6 cols=6 Triang Time=13;
     group=2 p rows=6 cols=6 Triang Time=19;
     design matrix constraints=24 covariates=24 identity;
   blabel(1)=Phi;
   blabel(2)=Phi;
   blabel(3)=Phi;
   blabel(4)=Phi;
   blabel(5)=Phi;
   blabel(6)=Phi;
   blabel(7)=Phi;
   blabel(8)=Phi;
   blabel(9)=Phi;
   blabel(10)=Phi;
   blabel(11)=Phi;
   blabel(12)=Phi;
   blabel(13)=p;
   blabel(14)=p;
   blabel(15)=p;
   blabel(16)=p;
   blabel(17)=p;
   blabel(18)=p;
   blabel(19)=p;
   blabel(20)=p;
   blabel(21)=p;
   blabel(22)=p;
   blabel(23)=p;
   blabel(24)=p;
   rlabel(1)=Phi;
   rlabel(2)=Phi;
   rlabel(3)=Phi;
   rlabel(4)=Phi;
   rlabel(5)=Phi;
   rlabel(6)=Phi;
   rlabel(7)=Phi;
   rlabel(8)=Phi;
   rlabel(9)=Phi;
   rlabel(10)=Phi;
   rlabel(11)=Phi;
   rlabel(12)=Phi;
   rlabel(13)=p;
   rlabel(14)=p;
   rlabel(15)=p;
   rlabel(16)=p;
   rlabel(17)=p;
   rlabel(18)=p;
   rlabel(19)=p;
   rlabel(20)=p;
   rlabel(21)=p;
   rlabel(22)=p;
   rlabel(23)=p;
   rlabel(24)=p;
   Hyperdistributions means=2 sigmas=2 rhos=1 vcmatrix;
      beta(13):0.0 1.75;
      beta(14):0.0 1.75;
      beta(15):0.0 1.75;
      beta(16):0.0 1.75;
      beta(17):0.0 1.75;
      beta(18):0.0 1.75;
      beta(19):0.0 1.75;
      beta(20):0.0 1.75;
      beta(21):0.0 1.75;
      beta(22):0.0 1.75;
      beta(23):0.0 1.75;
      beta(24):0.0 1.75;
      beta(1):mu(1) sigma(1);
      beta(2):mu(1) sigma(1);
      beta(3):mu(1) sigma(1);
      beta(4):mu(1) sigma(1);
      beta(5):mu(1) sigma(1);
      beta(6):mu(1) sigma(1);
      beta(7):mu(2) sigma(2);
      beta(8):mu(2) sigma(2);
      beta(9):mu(2) sigma(2);
      beta(10):mu(2) sigma(2);
      beta(11):mu(2) sigma(2);
      beta(12):mu(2) sigma(2);
      mu(1):0.4 Compute 0.0 100.0;
      mu(2):0.4 Compute 0.0 100.0;
      sigma(1):0.4 Compute 0.001 0.001;
      sigma(2):0.4 Compute 0.001 0.001;
      rho(1):0.4 0.0 -1.0 1.0;
Hypercorrelation rows=12;
sigma(1) 0 0 0 0 0 rho(1) 0 0 0 0 0;
0 sigma(1) 0 0 0 0 0 rho(1) 0 0 0 0;
0 0 sigma(1) 0 0 0 0 0 rho(1) 0 0 0;
0 0 0 sigma(1) 0 0 0 0 0 rho(1) 0 0;
0 0 0 0 sigma(1) 0 0 0 0 0 rho(1) 0;
0 0 0 0 0 sigma(1) 0 0 0 0 0 rho(1);
rho(1) 0 0 0 0 0 sigma(2) 0 0 0 0 0;
0 rho(1) 0 0 0 0 0 sigma(2) 0 0 0 0;
0 0 rho(1) 0 0 0 0 0 sigma(2) 0 0 0;
0 0 0 rho(1) 0 0 0 0 0 sigma(2) 0 0;
0 0 0 0 rho(1) 0 0 0 0 0 sigma(2) 0;
0 0 0 0 0 rho(1) 0 0 0 0 0 sigma(2);
proc stop;


FWIW, I just set up the Linux 64 bit on a updated Ubuntu and Mint, no issues thus far.

\bret

Hey, thats great. I was less worried about the Linux version, but glad to know its running under a couple of different distros. Thanks for the update.

[cooch]

Managed to arm-twist one of my students who had a newer, 64-bit Mac, and a few moments later, we now have a 64-bit version of MARK for OS/X.

Reminder that this is MARK from the command line (CLI), so for most Mac-users, the preferred interface will be using RMark. I know that MARK works from CLI -- I can't speak to anyone's success (or failure) in getting it to play nice with RMark.

[jlaake]

If someone out there has tried or can try RMark with the mark.exe that Evan has created for the mac, please let me know if you were successful or if you have had any problems. If there are problems I'll need someone's help to work through it because I don't have a mac. It works fine on a Linux machine so I'd think it would be the same for the mac. I want to release a new version to cran in the next few weeks so it would be nice to have any mac problems worked out for that release.

regards --jeff

[arpat]

Thanks a lot for compiling MARK for OSX. It saves the Mac-users from major headache, and I hope, not at the expense of more headache for you and Jeff.

So far it has been working fine for me (thanks to Evan and Jeff's input). My steps for getting it to work were:

- Install package RMark from source (not binaries) - check the box for "install dependencies"
- Download, rename and copy the "mark" file into /usr/local/bin. To open this hidden folder, type "open -a Finder /usr/local/bin" in Terminal. For some reason, any other directory (including the working directory) didn't work for me.
- In Terminal, check if the file is executable. Just typing "mark" and getting "No input file was specified, so MARK job is done." message verifies it. (If you are using an inp file to test, make sure the file includes the full set of MARK commands.)
- If you get an error (something like "dyld: lazy symbol binding failed...") install gfortran from http://gcc.gnu.org/wiki/GFortranBinaries#MacOS

cheers,
Arpat

[cooch]

In Terminal, check if the file is executable. Just typing "mark" and getting "No input file was specified, so MARK job is done." message verifies it. (If you are using an inp file to test, make sure the file includes the full set of MARK commands.)

An example of which was, in fact, posted earlier in this thread.