Installing Dalton on Ubuntu

[RickV]

Is there someone who can guide me to install Dalton on Ubuntu?

[arnfinn]

If you have your compilers installed (for instance gfortran and gcc), it should be rather straight forward.

Code: Select all

cd dalton
./setup
cd build
make

Please let us know if you are getting any problems or if you need more instructions.

[RickV]

The problem is probably that I don't have the compilers installed. But how do I get them?

[bast]

this installs compilers and math libs:

Code: Select all

sudo apt-get install gfortran gcc g++ libblas3gf libblas-dev liblapack3gf liblapack-dev

if you want MPI, you need to install either OpenMPI or MPICH.

[RickV]

Did the above. When I run ./setup, it tells me CMake is not found.

[bast]

you get cmake with:
sudo apt-get install cmake
this is also in our documentation:
http://daltonprogram.org/installation/2 ... or-too-old

[RickV]

Now, if I want to run the tests, I'm just getting the list of the tests with behind it (BAD_COMMAND) and at the end:
Errors while running CTest
make: *** [test] Error 8

* I'm running ctest in the build folder.

[RickV]

Example of my output for the last tests (i'm not copying the full list):
501 - LSint/LSDALTON_LSlib (BAD_COMMAND)
502 - LSresponse/LSresponse_HF_alpha (BAD_COMMAND)
503 - LSresponse/LSresponse_HF_alpha_aK (BAD_COMMAND)
504 - LSresponse/LSresponse_HF_alpha_astv (BAD_COMMAND)
505 - LSresponse/LSresponse_HF_beta (BAD_COMMAND)
506 - LSresponse/LSresponse_HF_dipole (BAD_COMMAND)
507 - LSresponse/LSresponse_HF_dtpa (BAD_COMMAND)
508 - LSresponse/LSresponse_HF_esd (BAD_COMMAND)
509 - LSresponse/LSresponse_HF_esg (BAD_COMMAND)
510 - LSresponse/LSresponse_HF_gamma (BAD_COMMAND)
511 - LSresponse/LSresponse_HF_gamma_lcm (BAD_COMMAND)
512 - LSresponse/LSresponse_HF_molgra (BAD_COMMAND)
513 - LSresponse/LSresponse_HF_opa (BAD_COMMAND)
514 - LSresponse/LSresponse_HF_tpa (BAD_COMMAND)
515 - LSresponse/LSresponse_DFT_alpha (BAD_COMMAND)
516 - LSresponse/LSresponse_DFT_alpha_aptv (BAD_COMMAND)
517 - LSresponse/LSresponse_DFT_beta (BAD_COMMAND)
518 - LSresponse/LSresponse_DFT_dipole (BAD_COMMAND)
519 - LSresponse/LSresponse_DFT_esd (BAD_COMMAND)
520 - LSresponse/LSresponse_DFT_esg (BAD_COMMAND)
521 - LSresponse/LSresponse_DFT_esg_LDA (BAD_COMMAND)
522 - LSresponse/LSresponse_DFT_excit (BAD_COMMAND)
523 - LSresponse/LSresponse_DFT_dtpa (BAD_COMMAND)
524 - LSresponse/LSresponse_DFT_molgra (BAD_COMMAND)
525 - LSresponse/LSresponse_DFT_opa (BAD_COMMAND)
526 - LSresponse/LSresponse_DFT_tpa (BAD_COMMAND)
527 - LSresponse/LSresponse_NMR_SHIELD (BAD_COMMAND)
528 - LSresponse/LSresponse_mcd (BAD_COMMAND)
529 - LSresponse/LSresponse_mcd2 (BAD_COMMAND)
530 - LSresponse/LSresponse_mcd3 (BAD_COMMAND)
531 - LSresponse/LSresponse_mcd_calc (BAD_COMMAND)
532 - LSresponse/LSresponse_excit_cs00 (BAD_COMMAND)
533 - geomopt/Excited_state_opt (BAD_COMMAND)
534 - geomopt/geoopt_constrain (BAD_COMMAND)
535 - geomopt/cartes_bfgs_min (BAD_COMMAND)
536 - geomopt/redint_bfgs (BAD_COMMAND)
537 - geomopt/cartes_trilevel_newconv (BAD_COMMAND)
538 - geomopt/cartes_psb_min (BAD_COMMAND)
539 - ddynam/HCN_ddyn (BAD_COMMAND)
540 - ddynam/HF_vib (BAD_COMMAND)
541 - ddynam/H2O_fmd (BAD_COMMAND)
542 - dectests/decmp2_energy (BAD_COMMAND)
543 - dectests/decmp2_density (BAD_COMMAND)
544 - dectests/decmp2_gradient (BAD_COMMAND)
545 - dectests/decmp2_geoopt (BAD_COMMAND)
546 - dectests/fullmp2_energy (BAD_COMMAND)
547 - dectests/fullcc2 (BAD_COMMAND)
548 - dectests/fullccsd_high (BAD_COMMAND)
549 - dectests/reordertest (BAD_COMMAND)
550 - dectests/decmp2_gradient_debug (BAD_COMMAND)
551 - plt/plt_dens (BAD_COMMAND)
552 - plt/plt_ep (BAD_COMMAND)
Errors while running CTest

[bast]

hi Rick,
haven't seen this before.
what i would do now is to run one single test (for example the dft_pbe test but any failing one will do):

Code: Select all

ctest -R dft_pbe

then i would look in build/Testing/Temporary/LastTest.log
for clues why it failed.
good luck!
radovan

[RickV]

What I get:
rick@ubuntu:~/Bureaublad/dalton/build$ ctest -R dft_pbe
Test project /home/rick/Bureaublad/dalton/build
Start 24: dft_pbe
1/1 Test #24: dft_pbe ..........................***Not Run 0.00 sec

0% tests passed, 1 tests failed out of 1

Label Time Summary:
dalton = 0.00 sec
dft = 0.00 sec
parallel = 0.00 sec
short = 0.00 sec

Total Test time (real) = 0.26 sec

The following tests FAILED:
24 - dft_pbe (BAD_COMMAND)
Errors while running CTest

[RickV]

I've started to recompile it again, see if something went wrong during compiling.

[bast]

please then try to run a test directly (ctest is a front-end to the TEST script):

Code: Select all

cd ~/Bureaublad/dalton/DALTON/test
./TEST -dalton ~/Bureaublad/dalton/build/dalton dft_pbe

[RickV]

That worked.

* rick@ubuntu:~/Bureaublad/dalton/DALTON/test$ ./TEST -dalton ~/Bureaublad/dalton/build/dalton dft_pbe

DALTON test : dft_pbe
invoked with : ./TEST -dalton /home/rick/Bureaublad/dalton/build/dalton dft_pbe
date and time : wo mrt 26 15:58:07 CET 2014
dalton script : /usr/bin/time /home/rick/Bureaublad/dalton/build/dalton
parameter list passed :
test job directory : 2014-03-26T15_58-testjob-pid-7933
test listing : DALTON_TESTLOG

dft_pbe
--------
Molecule: He
Wave Function: DFT PBE / cc-pVDZ
Test Purpose: PBE functional test.

0.38user 0.12system 0:00.50elapsed 101%CPU (0avgtext+0avgdata 34512maxresident)k
32inputs+89536outputs (1major+14638minor)pagefaults 0swaps
found /tmp and DALTON_TMPDIR set to /tmp
DALTON release 2013.2
Invocation: /home/rick/Bureaublad/dalton/build/dalton -d -ext log dft_pbe dft_pbe
wo mrt 26 15:58:07 CET 2014
Calculation: dft_pbe (input files: dft_pbe.dal and dft_pbe.mol)
Input dir : /home/rick/Bureaublad/dalton/DALTON/test/2014-03-26T15_58-testjob-pid-7933
Scratch dir: /tmp/DALTON_scratch_rick/dft_pbe_7962
DALTON: default work memory size used. 64000000
Work memory size (LMWORK+2): 64000002 = 488.28 megabytes; node 0
0: Directories for basis set searches:
/home/rick/Bureaublad/dalton/DALTON/test/2014-03-26T15_58-testjob-pid-7933:/home/rick/Bureaublad/dalton/build/basis
dft_pbe.tar.gz has been copied to /home/rick/Bureaublad/dalton/DALTON/test/2014-03-26T15_58-testjob-pid-7933
wo mrt 26 15:58:08 CET 2014

evaluate output file ./dft_pbe.log:
TEST ENDED PROPERLY -- 0 tests failed out of the 1 tests run so far

#####################################################################
Summary
#####################################################################

ALL TESTS ENDED PROPERLY!
date and time : wo mrt 26 15:58:08 CET 2014

[RickV]

ok, it's solved.
make test is running now.

Thanks for all!

[RickV]

Tests are running, but quite a lot of tests are failing. In the log files, I can find that the reason for failing is: Non-existing basis set in HERBAS.

Full test is still running (now at 115), but it failed at:
37
38
39
56
57
64
67

in the meantime also following tests failed:
117
118
119
123
157
158
159
160
166
171
172
173
175
176
177
178
179
181
182
184
185
189

[RickV]

Found the problem. For some reason, basis sets with * were named with _
e.g.: 6-31G* was named 6-31G_
Don't know why this was. Renaming the problem files solved it.

[bast]

interesting. so in dalton/basis they were with "*" and in dalton/build/basis with "_"?
or both with "_"?
the dalton/build/basis is created when you run setup.

[RickV]

Both dalton/basis and dalton/build/basis were with "_".
Though, now I only changed the _ to * in the build folder, which works fine now.

[bast]

how did you extract the tar.gz file?
the problem is that files with "*" in them are really problematic.
perhaps the tool that you used to extract the file renamed the files
for you?

[RickV]

I used winrar on windows, than i transferred the extracted folder with usb to a computer with ubuntu.

[bast]

ok next time you can download directly to linux and extract with:

Code: Select all

tar xvzf DALTON-2013.2-Source.tar.gz

windows does not allow "*" in file names.

[varuniitr]

I am installing DALTON by
./setup --mpi
Error is coming. Can anyone please tell how to get rid off this?Thanks in advance

/home/user/MPI/DALTON-2013.2-Source/DALTON/abacus/her2fck.F:514.23:

& offset_mat,
1
Error: Type mismatch in argument 'idispl' at (1); passed INTEGER(8) to INTEGER(4)
make[2]: *** [CMakeFiles/dalton.dir/DALTON/abacus/her2fck.F.o] Error 1
make[1]: *** [CMakeFiles/dalton.dir/all] Error 2
make: *** [all] Error 2

[bast]

dear Varun,
it seems that the same problem is discussed here:
http://daltonprogram.org/forum/viewtopi ... t_ma#p1203
have you installed the MPI library yourself? if so, how?
best wishes,
radovan

[varuniitr]

Sir I installed it using
sudo apt-get install libcr-dev mpich2 mpich2-doc.

[bast]

i think this is a bug in the mpich shipped with your version of Ubuntu.
i assume it's 1.4.1. latest mpich release is 3.1.1 which i recommend to
either compile from sources or you use precompiled 3.1.1 deb packages.
alternatively you can compile openmpi.