---- downloading matlab toolboxes ----

1) create a directory called `iap-idma' and two subdirectories
   called `iap-idma/matlab' and `iap-idma/release1'

2) open a web browser and download the following two files:
     ftp://mit.ecco-group.org/ecco_for_las/version_4/checkpoints/gcmfaces_20160114.tar.gz
     ftp://mit.ecco-group.org/ecco_for_las/version_4/checkpoints/c65r_MITprof.tar.gz

3) move these two files to the idma directory and uncompress them.

4) the iap-idma directory should now contain two new subdirectories
   named `gcmfaces' and `c65r_MITprof'. Rename c65r_MITprof as `MITprof'.
   At this point the iap-idma directory should contain 'matlab/',
   'release1/','MITprof/', and 'gmcfaces/'.

notes : 
- disk space requirement ~ 9M (plus session specific items that will be added later)
- 2011 or later matlab version is needed (for native netcdf implementation)
- for session 1 you only need MITprof but gcmfaces will come 
  handy later so your might as well download both right away

---- for session #1 ----

1) see 'downloading matlab toolboxes'

2) download idma_float_plot.m, idma_box_subset.m, and idma_box_mean.m from 
     http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/gael/comm/course-idma2016/matlab/

3) move these files to the iap-idma/matlab directory   

4) using you favorite ftp client (e.g. fetch in iOS, secureFX in MS-windows, 
   or wget in LINUX) to download the following directory:
      ftp://mit.ecco-group.org/ecco_for_las/version_4/release1/MITprof_release1
   
5) move MITprof_release1 to the iap-idma/release1 directory

6) start matlab, go to the iap-idma directory, and type:
   addpath matlab/
   [p]=idma_float_plot('4900828');
   box_l=[-180 -140]; box_L=[-10 10]-30; box_D=[0 300];
   [prof_T,prof_Testim,prof_Tweight]=idma_box_subset(box_l,box_L,box_D);
   [boxmean,tim,dep]=idma_box_mean(prof_T,prof_Tweight);
   figureL; pcolor(tim,-dep,boxmean); colorbar;

notes :
- disk space requirement ~ 6.4G
- idma_float_plot.m and idma_box_subset.m both add the 
  iap-idma/MITprof/ and its subdirectories to the matlab path
- idma_box_subset is the most time consuming of the three
  scripts and may take ~3 minutes

----- for session #2 -----

1) see 'downloading matlab toolboxes'
   note: make sure to use gcmfaces_20160114.tar.gz or the up to date version of
         gcmfaces from CVS (rather than c65r_MITprof.tar.gz or earlier versions).
2) download the m_map plotting toolbox (use the 'zip archive' link) from 
      https://www.eoas.ubc.ca/~rich/map.html 
   unzip and move it to iap-idma/
2) using you favorite ftp client (e.g. fetch in iOS, secureFX in MS-windows,
   or wget in LINUX) to download the following directory (145M):
      ftp://mit.ecco-group.org/ecco_for_las/version_4/release1/nctiles_grid
   and move it to iap-idma/
3) using you favorite ftp client (e.g. fetch in iOS, secureFX in MS-windows,
   or wget in LINUX) to download the following directory (14G):
      ftp://mit.ecco-group.org/ecco_for_las/version_4/release1/nctiles_climatology
   and move it to iap-idma/release1/
4) start matlab, go to the iap-idma directory, and type:
      p = genpath([pwd filesep 'gcmfaces/']); addpath(p);
      p = genpath([pwd filesep 'MITprof/']); addpath(p);
      p = genpath([pwd filesep 'm_map/']); addpath(p); 
      grid_load;%load ECCO v4 grid into mygrid
      gcmfaces_global;%declare mygrid as global
      disp(mygrid);%displays grid variables
      diags=example_transports;%computes transports (takes ~ 2 minutes)
      example_transports_disp(diags);%display results 
   then proceed as explained in section 5 of gcmfaces/gcmfaces.pdf
   to use diags_driver.m and diags_display.m (takes ~ 5 minutes)