fm2dstt5.f90: test to see how sparsity related to smoothness, no sparsity first fm2dstt5_test.f90 sparsity constrain, only to the model variation fm2dstt6.f90 fm2dstt5_test.f90 with scaling fm2dstt7.f90 add ray geometry to file fm2dstt8.f90 add openmp fm2dstt8_2.f90 a little change,seems bug fm2dstt9.f90 add weight to the data residual fm2dstt10.f90 l1 norm minimization to decrease the outlier effect the result seems not so good fm2dstt10_2.f90 change to another verison, incorporate the regularization term into the misfit term fm2dstt10_3.f90 add the scaling relationship of Vp and rho into the sensitivity matrix fm2dstt10_4.f90 probably a bug in the weight matrix of IRLS, use sqrt to the weight fm2dstt11.f90 fm2dstt8.f90 without scaling,the result seems good compared to the one with scaling, with smaller weight... fm2dstt11_2.f90 add the scaling relationship of Vp and rho into the sensitivity matrix --------------------------------- SurfTomo3.2 new format of input file like joint inversion code #remember to renew the format again in a format that computing the traveltime for one source and then get the raypath for all receivers, this will save some time since computing the travel time field every time is unnecessary