static void *
run_clip_permutation(void *args)
{
CLIPSTATS_TREES *trees;
JOB_COUNTER *jobcounter;
RANDGEN_STATE randgen;
WORKER *worker;
gzFile fp=NULL;
randgen_init(&randgen);
worker = (WORKER *)args;
jobcounter = worker->jobcounter;
fp = gzopen(worker->readpool_input_path, "r");
if (fp == NULL) {
fprintf(stderr, "Failed to open input file %s\n", worker->readpool_input_path);
return NULL;
}
trees = clipstatstrees_new();
#define BEGIN_JOBCOUNTER_EXCLUSIVE pthread_mutex_lock(&jobcounter->lock);
#define END_JOBCOUNTER_EXCLUSIVE pthread_mutex_unlock(&jobcounter->lock);
#define LOCK_AND_UPDATE_STATUS(worker, newstatus) do { \
pthread_mutex_lock(&(worker)->jobcounter->lock); \
(worker)->status = newstatus; \
update_progression((worker)); \
pthread_mutex_unlock(&(worker)->jobcounter->lock); \
} while (0)
for (;;) {
READ_QUEUE_SET *read_queue_set=NULL;
struct timeval tv;
/* Increase the job counter, exit the loop if all works done */
BEGIN_JOBCOUNTER_EXCLUSIVE
if (jobcounter->queued >= jobcounter->total) {
END_JOBCOUNTER_EXCLUSIVE
break;
}
jobcounter->queued++;
gettimeofday(&worker->last_started, NULL);
worker->status = WORKER_STATUS_READ_QUEUE_SHUFFLING;
update_progression(worker);
END_JOBCOUNTER_EXCLUSIVE
/* Generate shuffled read queue from error profile */
read_queue_set = generate_read_queue_set(&randgen,
worker->error_profile);
if (read_queue_set == NULL) {
LOCK_AND_UPDATE_STATUS(worker, WORKER_STATUS_ERROR);
fclose(fp);
return NULL;
}
/* Run simulated sequencing and get profiles */
LOCK_AND_UPDATE_STATUS(worker, WORKER_STATUS_SIMULATING);
gzrewind(fp);
simulate_sequencing(fp, read_queue_set, trees);
free_read_queue_set(read_queue_set);
/* Increase the job counter for jobs done */
BEGIN_JOBCOUNTER_EXCLUSIVE
jobcounter->done++;
worker->status = WORKER_STATUS_NOT_RUNNING;
gettimeofday(&tv, NULL);
worker->last_consumed = TIMEVAL_DIFF(worker->last_started, tv);
update_progression(worker);
END_JOBCOUNTER_EXCLUSIVE
}
int main(void)
{
mat *a,*b,*c,*d,*e,*f,*g,*h,*i,*j;
const double b_init[] =
{
2.5,\
2.0,\
5.0
};
const double c_init[] =
{
1.0,10.0,6.7
};
a = mat_create(3,3);
b = mat_create_from_ar(b_init,3,1);
c = mat_create_from_ar(c_init,1,3);
d = mat_create(2,2);
e = mat_create(3,3);
f = mat_create(3,3);
g = mat_create(5,3);
h = mat_create(3,5);
i = mat_create(5,3);
j = mat_create(3,1);
randgen_init(12);
latan_set_verb(DEBUG1);
printf("---------- %-30s ----------\n","matrix product");
printf("b =\n");
mat_print(b,"%8.2f");
printf("\n");
printf("c =\n");
mat_print(c,"%8.2f");
printf("\n");
printf("a <- b*c\n");
mat_mul(a,b,'n',c,'n');
printf("a =\n");
mat_print(a,"%8.2f");
printf("\n");
printf("\n---------- %-30s ----------\n","sub-matrices");
printf("d <- a(1:2,0:1)\n");
mat_get_subm(d,a,1,0,2,1);
printf("d =\n");
mat_print(d,"%9.2f");
printf("\n");
printf("a(1,0:2) <- c\n");
mat_set_subm(a,c,1,0,1,2);
printf("a =\n");
mat_print(a,"%8.2f");
printf("\n");
printf("\n---------- %-30s ----------\n","symmetric matrix product");
printf("e <- a*t(a)\n");
mat_mul(e,a,'n',a,'t');
mat_print(e,"%8.2f");
printf("\n");
printf("e is assumed symmetric\n\n");
mat_assume(e,MAT_SYM);
printf("j <- e*b\n");
mat_mul(j,e,'n',b,'t');
mat_print(j,"%8.2f");
printf("\n");
printf("f <- a*e\n");
mat_mul(f,a,'n',e,'n');
mat_print(f,"%8.2f");
printf("\n");
printf("f <- e*t(a)\n");
mat_mul(f,e,'n',a,'t');
mat_print(f,"%8.2f");
printf("\n");
printf("g <- random(-1.0,1.0)\n");
mat_rand_u(g,-1.0,1.0);
mat_print(g,"%8.2f");
printf("\n");
printf("h <- e*t(g) (buffer should be created)\n");
mat_mul(h,e,'n',g,'t');
mat_print(h,"%8.2f");
printf("\n");
printf("i <- g*e\n");
mat_mul(i,g,'n',e,'n');
mat_print(i,"%8.2f");
printf("\n");
printf("e is not assumed symmetric anymore\n");
mat_reset_assump(e);
printf("\n---------- %-30s ----------\n","matrix inversion");
printf("*** LU decomposition\n");
printf("a <- random(-6.0,6.0)\n");
mat_rand_u(a,-6.0,6.0);
printf("a =\n");
mat_print(a,"%8.2f");
printf("\n");
printf("e <- a^(-1)\n");
mat_inv_LU(e,a);
printf("e =\n");
mat_print(e,"%8.2f");
printf("\n");
printf("f <- e*a\n");
mat_mul(f,e,'n',a,'n');
printf("f =\n");
mat_print(f,"%8.2f");
printf("\n");
printf("*** pseudo-inverse with good-conditioned matrix\n");
printf("a =\n");
mat_print(a,"%8.2f");
printf("\n");
printf("e <- a^+\n");
mat_pseudoinv(e,a);
printf("e =\n");
mat_print(e,"%8.2f");
printf("\n");
printf("f <- e*a\n");
mat_mul(f,e,'n',a,'n');
printf("f =\n");
mat_print(f,"%8.2f");
printf("\n");
printf("*** symmetric pseudo-inverse with good-conditioned matrix\n");
printf("a <- a*t(a)\n");
mat_mul(a,a,'n',a,'t');
mat_print(a,"%8.2f");
printf("\n");
printf("a is assumed symmetric and positive\n\n");
mat_assume(a,(mat_flag)(MAT_SYM|MAT_POS));
printf("e <- a^+\n");
mat_pseudoinv(e,a);
printf("e =\n");
mat_print(e,"%8.2f");
printf("\n");
printf("f <- e*a\n");
mat_mul(f,e,'n',a,'n');
printf("f =\n");
mat_print(f,"%8.2f");
printf("\n");
printf("*** Cholesky decomposition\n");
printf("e <- a^(-1)\n");
mat_inv_symChol(e,a);
printf("e =\n");
mat_print(e,"%8.2f");
printf("\n");
printf("f <- e*a\n");
mat_mul(f,e,'n',a,'n');
printf("f =\n");
mat_print(f,"%8.2f");
printf("\n");
mat_destroy(a);
mat_destroy(b);
mat_destroy(c);
mat_destroy(d);
mat_destroy(e);
mat_destroy(f);
mat_destroy(g);
mat_destroy(h);
mat_destroy(i);
mat_destroy(j);
return EXIT_SUCCESS;
}
/*
**
** void start_randgen(int iSeed1, int iSeed2)
**
** Seeds the random number generator
**
*/
void start_randgen(int iSeed1, int iSeed2)
{
randgen_init(iSeed1, iSeed2);
}
