popgenmethods / smcpp

SMC++ infers population history from whole-genome sequence data.
GNU General Public License v3.0
149 stars 34 forks source link

error with new version 1.15.4 #236

Open ndussex opened 1 year ago

ndussex commented 1 year ago

Hi,

I was trying to run the new version 1.15.4 hoping it would not give me error messages regarding missing data mask files, but I now seem to have an issue with the format of my files when I run the smc++ estimate of cv function.

Traceback (most recent call last): File "/sw/bioinfo/SMC++/1.15.4/rackham/bin/smc++", line 8, in sys.exit(main()) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/frontend/console.py", line 28, in main cmds[args.command].main(args) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/commands/estimate.py", line 47, in main analysis = Analysis(args.data, args) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/analysis/analysis.py", line 57, in init self.run(1) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/analysis/base.py", line 121, in run self._optimizer.run(niter or self._niter) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/optimize/optimizers.py", line 176, in run res = self._minimize(x0, coords) File "/sw/bioinfo/SMC++/1.15.4/rackham/venv/lib/python3.10/site-packages/smcpp/optimize/optimizers.py", line 108, in _minimize res = scipy.optimize.minimize(self._f, x0z, File "/sw/comp/python/3.10.8/rackham/lib/python3.10/site-packages/scipy/optimize/_minimize.py", line 690, in minimize res = _minimize_powell(fun, x0, args, callback, bounds, **options) File "/sw/comp/python/3.10.8/rackham/lib/python3.10/site-packages/scipy/optimize/_optimize.py", line 3254, in _minimize_powell elif np.isnan(fval) or np.isnan(x).any(): TypeError: ufunc 'isnan' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''

The snc++ vcf2scm seems to be working without any error. Has anyone encountered this issue before?

Regards, Nic