DMFT: Input/output files

From ALPS
Jump to: navigation, search

The files with prefix BASENAME: (where BASENAME is the name of the parameter input file)

  • BASENAME: it is the input file to be loaded by the application dmft
  • BASENAME.h5: contains the iteration resolved impurity Green's function G(\tau)and the Weiss field G^0(\tau) in the imaginary time representation; if the selfconsistency loop has been performed in Matsubara representation (= if OMEGA_LOOP has been on) then there will be stored the G(i\omega_n) and G^0(i\omega_n) as well. The selfenergy is there not stored directly, but may be obtained via Dyson equation easily (look into DMFT-01 An introduction to DMFT)


The output/input files in Matsubara representation: (text file which consists of NMATSUBARA rows, each for one Matsubara frequency)

  • G_omega_i (G0_omega_i): contains the imaginary part of the Green's function (Weiss field) given in Matsubara frequencies after the i-th iteration; rows contain the \omega_n followed by the imaginary part of the Green's function (Weiss field) for each flavor; thus there are 1+FLAVORS columns in the file
  • G_omegareal_i (G0_omegareal_i): the same as above for the real part
  • selfenergy_i: contains the selfenergy after the i-th iteration; each row consists of \omega_n followed by the real and imaginary part of the selfenergy for each flavor; thus there are 1+2FLAVORS columns in the file
  • G0omega_output (unless not specified differently by the variable G0OMEGA_output): contains the n (corresponding to \omega_n=\frac{(2n+1)\pi}{\beta}) followed by the complex Weiss field for each flavor; thus there is one integer column followed by FLAVORS columns of complex numbers defined by the real and imaginary part in brackets
  • G0OMEGA_INPUT: variable specifying the input file with the initial Weiss field in Matsubara representation; does expect the same format as the above output file; thus you may copy it and start a simulation from it


The output/input files in imaginary time representation: (text file which consists of N+1 rows, each for one imaginary time \in<0,\beta>)

  • G_tau_i (G0_tau_i): contains the (real) Green's function (Weiss field) after the i-th iteration; rows contain the \tau_n followed by the Green's function (Weiss field) for each flavor; thus there are 1+FLAVORS columns in the file
  • G0tau_output (unless not specified differently by the variable G0TAU_output): contains the n (corresponding to \tau_n=\frac{n}{N}\beta) followed by the complex Weiss field for each flavor; thus there is one integer column followed by FLAVORS columns of complex numbers defined by the real and imaginary part in brackets; in total N+1 rows
  • G0OMEGA_INPUT: variable specifying the input file with the initial Weiss field in imaginary time representation; does expect the same format as the above output file; thus you may copy it and start a simulation from it


The output files with prefix given by the optional variable CHECKPOINT:

  • CHECKPOINT.h5: contains the measurements for each iteration
  • CHECKPOINT.xml: contains the input parameters and run information
  • CHECKPOINT.run*: contains information to rerun the simulation (these are the true checkpoints); for each process


The output files for the hybridization expansion impurity solver: (text files)

  • overlap: i-th row contains the <n_\downarrow n_\uparrow> in the i-th iteration
  • matrix_size: