-lmm 9 option leads to bug in mvlmm model

[diogro]

The mvlmm file is missing the case for the -lmm 9 option, leading to an output file with a single line and no significance values.

GEMMA/src/mvlmm.cpp

Lines 192 to 202 in 5caf948

	if (a_mode == 1) {
	outfile << sumStat[t].p_wald << endl;
	} else if (a_mode == 2) {
	outfile << sumStat[t].p_lrt << endl;
	} else if (a_mode == 3) {
	outfile << sumStat[t].p_score << endl;
	} else if (a_mode == 4) {
	outfile << sumStat[t].p_wald << "\t" << sumStat[t].p_lrt << "\t"
	<< sumStat[t].p_score << endl;
	} else {
	}

I stumbled on this because I'm also experiencing the long run times referenced in issue #243 when using -lmm 1. Option 9 runs, but with the mangled output.

[pjotrp]

Thanks for reporting!