static void write_coarse
( mcxIO* xf_coarse
, mclx* mx_coarse
)
{ mclx* tmp = mclxCopy(mx_coarse)
; double mv = mclxMaxValue(tmp)
; double sc = mv ? mv / 100.0 : 1.0
; mclxUnary(tmp, fltxScale, &sc)
; mclxaWrite(tmp, xf_coarse, 4, RETURN_ON_FAIL)
; mclxFree(&tmp)
; }
static mclx* get_coarse
( const mclx* mxbase
, mclx* clprev
, mcxbool add_transpose
)
{ mclx* blockc = mclxBlocksC(mxbase, clprev)
; mclx* clprevtp = mclxTranspose(clprev)
; mclx *p1 = NULL /* p_roduct */
; mclx* mx_coarse= NULL
; mclxMakeStochastic(clprev)
/****************** <EXPERIMENTAL CRAP> ************************************/
; if (hdp_g)
mclxUnary(clprev, fltxPower, &hdp_g)
/* parameter: use mxbase rather than blockc */
; if (getenv("MCLCM_BLOCK_STOCHASTIC")) /* this works very badly! */
mclxMakeStochastic(blockc)
; else if (getenv("MCLCM_BASE_UNSCALE") && start_col_sums_g)
{ dim i
; for (i=0;i<N_COLS(blockc);i++)
{ double f = start_col_sums_g->ivps[i].val
; mclvUnary(blockc->cols+i, fltxMul, &f)
; }
; }
/****************** </EXPERIMENTAL> *****************************************/
p1 = mclxCompose(blockc, clprev, 0)
;if (0)
{mcxIO* t = mcxIOnew("-", "w")
;mclxWrite(blockc, t, MCLXIO_VALUE_GETENV, EXIT_ON_FAIL)
;
}
; mclxFree(&blockc)
; mx_coarse = mclxCompose(clprevtp, p1, 0)
; if (add_transpose)
mclxAddTranspose(mx_coarse, 0.0)
; mclxAdjustLoops(mx_coarse, mclxLoopCBremove, NULL)
; mclxFree(&p1)
; mclxFree(&clprevtp)
; mclxMakeCharacteristic(clprev)
; return mx_coarse
; }
int main
( int argc
, const char* argv[]
)
{ mcxIO* xfmx = mcxIOnew("-", "r"), *xfout = mcxIOnew("-", "w")
; mclx* mx = NULL
; mclv* mx_diag = NULL
; mcxstatus parseStatus = STATUS_OK
; mcxOption* opts, *opt
; dim N_edge = 0
; dim* offsets
; dim template_n_nodes = 0
; mcxbool plus = FALSE
; double e_min = 1.0
; double e_max = 0.0
; double skew = 0.0
; double radius = 0.0
; double n_sdev = 0.5
; double n_range = 2.0
; double g_radius = 0.0
; double g_mean = 0.0
; double g_sdev = 0.0
; double g_min = 1.0
; double g_max = 0.0
; mcxbool do_gaussian = FALSE
; dim i = 0
; dim N_remove = 0
; dim N_add = 0
; dim N_shuffle = 0
; unsigned long random_ignore = 0
; srandom(mcxSeed(2308947))
; mcxOptAnchorSortById(options, sizeof(options)/sizeof(mcxOptAnchor) -1)
; if
(!(opts = mcxOptParse(options, (char**) argv, argc, 1, 0, &parseStatus)))
exit(0)
; mcxLogLevel =
MCX_LOG_AGGR | MCX_LOG_MODULE | MCX_LOG_IO | MCX_LOG_GAUGE | MCX_LOG_WARN
; mclxIOsetQMode("MCLXIOVERBOSITY", MCL_APP_VB_YES)
; mclx_app_init(stderr)
; for (opt=opts;opt->anch;opt++)
{ mcxOptAnchor* anch = opt->anch
; switch(anch->id)
{ case MY_OPT_HELP
: case MY_OPT_APROPOS
: mcxOptApropos(stdout, me, syntax, 20, MCX_OPT_DISPLAY_SKIP, options)
; return 0
;
case MY_OPT_VERSION
: app_report_version(me)
; return 0
;
case MY_OPT_SKEW
: skew = atof(opt->val)
; break
;
case MY_OPT_GEN
: template_n_nodes = atoi(opt->val)
; break
;
case MY_OPT_IMX
: mcxIOrenew(xfmx, opt->val, NULL)
; break
;
case MY_OPT_PLUS
: case MY_OPT_WB
: plus = TRUE
; break
;
case MY_OPT_OUT
: mcxIOrenew(xfout, opt->val, NULL)
; break
;
case MY_OPT_E_MAX
: if (!strcmp(opt->val, "copy"))
e_max = -DBL_MAX
; else
e_max = atof(opt->val)
; break
;
case MY_OPT_E_MIN
: e_min = atof(opt->val)
; break
;
case MY_OPT_G_MIN
: g_min = atof(opt->val)
; break
;
case MY_OPT_G_MAX
: g_max = atof(opt->val)
; break
;
case MY_OPT_G_SDEV
: g_sdev = atof(opt->val)
; break
;
case MY_OPT_G_MEAN
: g_mean = atof(opt->val)
; do_gaussian = TRUE
; break
;
case MY_OPT_G_RADIUS
: g_radius = atof(opt->val)
; break
;
case MY_OPT_N_RANGE
: n_range = atof(opt->val)
; break
;
case MY_OPT_N_SDEV
: n_sdev = atof(opt->val)
; break
;
case MY_OPT_N_RADIUS
: radius = atof(opt->val)
; break
;
case MY_OPT_SHUFFLE
: N_shuffle = atoi(opt->val)
; break
;
case MY_OPT_ADD
: N_add = atoi(opt->val)
; break
;
case MY_OPT_REMOVE
: N_remove = atoi(opt->val)
; break
; }
}
/* hitting y% in vi tells me the size of this block */
{ if (template_n_nodes)
mx = mclxAllocZero
( mclvCanonical(NULL, template_n_nodes, 1.0)
, mclvCanonical(NULL, template_n_nodes, 1.0)
)
; else
mx = mclxReadx
( xfmx
, EXIT_ON_FAIL
, MCLX_REQUIRE_GRAPH
)
; mx_diag = mclxDiagValues(mx, MCL_VECTOR_COMPLETE)
; if (N_shuffle)
mclxAdjustLoops(mx, mclxLoopCBremove, NULL)
; else
mclxSelectUpper(mx)
/* ^ apparently we always work on single arc representation (docme andsoon) */
; offsets = mcxAlloc(sizeof offsets[0] * N_COLS(mx), EXIT_ON_FAIL)
; N_edge = 0
; for (i=0;i<N_COLS(mx);i++)
{ offsets[i] = N_edge
; N_edge += mx->cols[i].n_ivps
; }
if (N_edge < N_remove)
{ mcxErr
( me
, "removal count %ld exceeds edge count %ld"
, (long) N_remove
, (long) N_edge
)
; N_remove = N_edge
; }
random_ignore = RAND_MAX - (N_edge ? RAND_MAX % N_edge : 0)
; if (RAND_MAX / 2 < N_edge)
mcxDie(1, me, "graph too large!")
; if (N_shuffle)
{ do_the_shuffle(mx, N_shuffle, offsets, N_edge, random_ignore)
; mx_readd_diagonal(mx, mx_diag)
; mclxWrite(mx, xfout, MCLXIO_VALUE_GETENV, RETURN_ON_FAIL)
; exit(0)
; }
; if (N_remove)
{ dim n_remove = do_remove(mx, N_remove, offsets, N_edge, random_ignore)
/* Need to recompute N_edge and random_ignore.
* NOTE we work with *upper* matrix; this counts graph edges.
*/
; N_edge = mclxNrofEntries(mx) - n_remove
; random_ignore = RAND_MAX - (RAND_MAX % N_COLS(mx))
; }
if (g_mean)
{ if (!g_radius)
{ if (g_sdev)
g_radius = 2 * g_sdev
; mcxWarn(me, "set radius to %.5f\n", g_radius)
; }
}
; if (N_add)
N_edge += do_add
( mx
, N_add , N_edge
, do_gaussian ? &g_mean : NULL, g_radius , g_sdev , g_min , g_max
, skew
, e_min , e_max
)
; if (radius)
{ for (i=0;i<N_COLS(mx);i++)
{ mclp* ivp = mx->cols[i].ivps, *ivpmax = ivp + mx->cols[i].n_ivps
;if(DEBUG)fprintf(stderr, "here %d\n", (int) i)
; while (ivp < ivpmax)
{ double val = ivp->val
; double r = mcxNormalCut(n_range * n_sdev, n_sdev)
; double newval = val + radius * (r / (n_range * n_sdev))
; if (e_min < e_max && newval >= e_min && newval <= e_max)
; ivp->val = newval
; ivp++
; }
}
}
mclxUnary(mx, fltxCopy, NULL) /* remove zeroes */
; mclxAddTranspose(mx, 0.0)
; mx_readd_diagonal(mx, mx_diag)
; if (plus)
mclxbWrite(mx, xfout, RETURN_ON_FAIL)
; else
mclxWrite(mx, xfout, MCLXIO_VALUE_GETENV, RETURN_ON_FAIL)
; }
return 0
; }
static dim clm_clm_prune
( mclx* mx
, mclx* cl
, dim prune_sz
, mclx** cl_adjustedpp
, dim* n_sink
, dim* n_source
)
{ dim d, n_adjusted = 0
; mclx* cl_adj = mclxCopy(cl)
; mclv* cid_affected = mclvClone(cl->dom_cols)
; const char* me = "clmAssimilate"
; double bar_affected = 1.5
; mclx *el_to_cl = NULL
; mclx *el_on_cl = NULL
; mclx *cl_on_cl = NULL
; mclx *cl_on_el = NULL
; *n_sink = 0
; *n_source = 0
; mclvMakeConstant(cid_affected, 1.0)
; mclxColumnsRealign(cl_adj, mclvSizeCmp)
; *cl_adjustedpp = NULL
; clmCastActors
(&mx, &cl_adj, &el_to_cl, &el_on_cl, &cl_on_cl, &cl_on_el, 0.95)
; mclxFree(&cl_on_el)
; for (d=0;d<N_COLS(cl_on_cl);d++)
{ mclv* clthis = cl_adj->cols+d
; mclv* cllist = cl_on_cl->cols+d
; mclp* pself = mclvGetIvp(cllist, clthis->vid, NULL)
; double self_val = -1.0
; if (pself)
self_val = pself->val
, pself->val *= 1.001 /* to push it up in case of equal weights */
;if(0)fprintf(stderr, "test size %d\n", (int) clthis->n_ivps)
; if (prune_sz && clthis->n_ivps > prune_sz)
continue
; while (1)
{ mclv* clthat
; dim e
; if (cllist->n_ivps < 2)
break
; mclvSort(cllist, mclpValRevCmp)
/* now get biggest mass provided that cluster
* ranks higher (has at least as many entries)
*
* fixme/todo: we probably have a slight order
* dependency for some fringe cases. If provable
* then either solve or document it.
*/
; for (e=0;e<cllist->n_ivps;e++)
if (cllist->ivps[e].idx >= clthis->vid)
break
/* found none or itself */
; if (e == cllist->n_ivps || cllist->ivps[e].idx == clthis->vid)
break
; if /* Should Not Happen */
(!(clthat
= mclxGetVector(cl_adj, cllist->ivps[e].idx, RETURN_ON_FAIL, NULL)
) )
break
/* works for special case prune_sz == 0 */
/* if (clthat->n_ivps + clthis->n_ivps > prune_sz) */
/* ^iced. inconsistent behaviour as k grows. */
; { mcxLog
( MCX_LOG_LIST
, me
, "source %ld|%lu|%.3f absorbed by %ld|%lu|%.3f"
, clthis->vid, (ulong) clthis->n_ivps, self_val
, clthat->vid, (ulong) clthat->n_ivps, cllist->ivps[0].val
)
; n_adjusted += clthis->n_ivps
; (*n_sink)++
/* note: we could from our precomputed cl_on_cl
* obtain that A is absorbed in B, B is absorbed in C.
* below we see that A will be merged with B,
* and the result will then be merged with C.
* This depends on the fact that cl_adj is ordered
* on increasing cluster size.
*/
; mcldMerge(cl_adj->cols+d, clthat, clthat)
; mclvResize(cl_adj->cols+d, 0)
; mclvInsertIdx(cid_affected, clthat->vid, 2.0)
; }
break
; }
mclvSort(cllist, mclpIdxCmp)
; }
mclxFree(&cl_on_cl)
; mclxFree(&el_on_cl)
; mclxFree(&el_to_cl)
; mclxMakeCharacteristic(cl)
; mclvUnary(cid_affected, fltxGT, &bar_affected)
; *n_source = cid_affected->n_ivps
; mclvFree(&cid_affected)
; mclxColumnsRealign(cl_adj, mclvSizeRevCmp)
; if (!n_adjusted)
{ mclxFree(&cl_adj)
; return 0
; }
mclxUnary(cl_adj, fltxCopy, NULL)
; mclxMakeCharacteristic(cl_adj)
; *cl_adjustedpp = cl_adj
; return n_adjusted
; }
static dim clm_clm_adjust
( mclx* mx
, mclx* cl
, dim cls_size_max
, mclx** cl_adjustedpp
, mclv** cid_affectedpp
, mclv** nid_affectedpp
)
{ dim i, j, n_adjusted = 0
; mclx* cl_adj = mclxCopy(cl)
; mclv* cid_affected = mclvClone(cl->dom_cols)
; mclv* nid_affected = mclvClone(mx->dom_cols)
; double bar_affected = 1.5
; const char* e1 = getenv("MCL_ADJ_FMAX")
; const char* e2 = getenv("MCL_ADJ_EMASS")
; double f1 = e1 ? atof(e1) : 2
; double f2 = e2 ? atof(e2) : 3
; mcxbool loggit = mcxLogGet( MCX_LOG_CELL | MCX_LOG_INFO )
; clmVScore sc
; mclx *el_to_cl = NULL
; mclx *el_on_cl = NULL
; mclx *cl_on_cl = NULL
; mclx *cl_on_el = NULL
; *cl_adjustedpp = NULL
; *cid_affectedpp = NULL
; *nid_affectedpp = NULL
; clmCastActors
(&mx, &cl, &el_to_cl, &el_on_cl, &cl_on_cl, &cl_on_el, 0.95)
; mclxFree(&cl_on_cl)
; mclxFree(&cl_on_el)
; mclvMakeConstant(cid_affected, 1.0)
; mclvMakeConstant(nid_affected, 1.0)
; for (i=0;i<N_COLS(cl_adj);i++)
cl_adj->cols[i].val = 0.5
/* Proceed with smallest clusters first.
* Caller has to take care of mclxColumnsRealign
*/
; for (i=0;i<N_COLS(cl);i++)
{ mclv* clself = cl->cols+i
/* Only consider nodes in clusters of
* size <= cls_size_max
*/
; if (cls_size_max && clself->n_ivps > cls_size_max)
break
/* Clusters that have been marked for inclusion
* cannot play.
*/
; if (cl_adj->cols[i].val > 1)
continue
; for (j=0;j<clself->n_ivps;j++)
{ long nid = clself->ivps[j].idx
; long nos = mclvGetIvpOffset(mx->dom_cols, nid, -1)
; mclv* clidvec = mclxGetVector(el_on_cl, nid, RETURN_ON_FAIL, NULL)
; double eff_alien_bsf = 0.0, eff_alien_max_bsf = 0.0 /* best so far*/
; double eff_self = 0.0, eff_self_max = 0.0
; long cid_alien = -1, cid_self = -1
; clmVScore sc_self = { 0 }, sc_alien = { 0 }
; dim f
; if (nos < 0 || !clidvec)
{ mcxErr
("clmDumpNodeScores panic", "node <%ld> does not belong", nid)
; continue
; }
clmVScanDomain(mx->cols+nos, clself, &sc)
; clmVScoreCoverage(&sc, &eff_self, &eff_self_max)
; cid_self = clself->vid
; sc_self = sc
; if (loggit)
mcxLog2
( us
, "node %ld in cluster %ld eff %.3f,%.3f sum %.3f"
, nid
, cid_self
, eff_self
, eff_self_max
, sc.sum_i
)
; for (f=0;f<clidvec->n_ivps;f++)
{ long cid = clidvec->ivps[f].idx
; mclv* clvec = mclxGetVector(cl, cid, RETURN_ON_FAIL, NULL)
/* ^ overdoing: cid == clvec->vid */
; double eff, eff_max
; if (!clvec)
{ mcxErr
( "clmAdjust panic"
, "cluster <%ld> node <%ld> mishap"
, cid
, nid
)
; continue
; }
/* fixme: document or remove first condition
*
*/
if ((0 && clvec->n_ivps <= clself->n_ivps) || clvec->vid == cid_self)
continue
; clmVScanDomain(mx->cols+nos, clvec, &sc)
; clmVScoreCoverage(&sc, &eff, &eff_max)
#if 0
# define PIVOT eff > eff_alien_bsf
#else
# define PIVOT eff_max > eff_alien_max_bsf
#endif
; if
( PIVOT
|| sc.sum_i >= 0.5
)
eff_alien_bsf = eff
, eff_alien_max_bsf = eff_max
, cid_alien = clvec->vid
, sc_alien = sc
; if (sc.sum_i >= 0.5)
break
; }
if (loggit)
mcxLog2
( us
, " -> best alien %ld eff %.3f,%.3f sum %.3f"
, cid_alien
, eff_alien_bsf
, eff_alien_max_bsf
, sc_alien.sum_i
)
/* below: use sum_i as mass fraction
* (clmAdjust framework uses stochastic * matrix)
*/
; if
( cid_alien >= 0
&& cid_self >= 0
&& f1 * sc_alien.max_i >= sc_self.max_i
&& ( ( eff_alien_bsf > eff_self
&& sc_alien.sum_i > sc_self.sum_i
)
|| ( pow(sc_alien.sum_i, f2) >= sc_self.sum_i
&& pow(eff_self, f2) <= eff_alien_bsf
)
)
/* So, if max is reasonable
* and efficiency is better and mass is better
* or if mass is ridiculously better -> move
* Somewhat intricate and contrived, yes.
*/
)
{ mclv* acceptor
= mclxGetVector(cl_adj, cid_alien, RETURN_ON_FAIL, NULL)
; mclv* donor
= mclxGetVector(cl_adj, cid_self, RETURN_ON_FAIL, NULL)
; if (!donor || !acceptor || acceptor == donor)
continue
; mclvInsertIdx(donor, nid, 0.0)
; mclvInsertIdx(acceptor, nid, 1.0)
; acceptor->val = 1.5
; if (mcxLogGet(MCX_LOG_LIST))
{ mclv* nb = mx->cols+nos
; double mxv = mclvMaxValue(nb)
; double avg = nb->n_ivps ? mclvSum(nb) / nb->n_ivps : -1.0
; mcxLog
( MCX_LOG_LIST
, us
, "mov %ld (%ld %.2f %.2f)"
" %ld (cv=%.2f cm=%.2f s=%.2f m=%.2f #=%lu)"
" to %ld (cv=%.2f cm=%.2f s=%.2f m=%.2f #=%lu)"
, nid
, (long) nb->n_ivps, mxv, avg
, cid_self
, eff_self, eff_self_max, sc_self.sum_i, sc_self.max_i
, (ulong) (sc_self.n_meet + sc_self.n_ddif)
, cid_alien
, eff_alien_bsf, eff_alien_max_bsf, sc_alien.sum_i, sc_alien.max_i
, (ulong) (sc_alien.n_meet + sc_alien.n_ddif)
)
; }
n_adjusted++
; mclvInsertIdx(cid_affected, cid_alien, 2.0)
; mclvInsertIdx(cid_affected, cid_self, 2.0)
; mclvInsertIdx(nid_affected, nid, 2.0)
; }
}
}
mclxFree(&el_on_cl)
; mclxFree(&el_to_cl)
; for (i=0;i<N_COLS(cl_adj);i++)
cl_adj->cols[i].val = 0.0
; mclxMakeCharacteristic(cl)
; if (!n_adjusted)
{ mclxFree(&cl_adj)
; mclvFree(&cid_affected)
; mclvFree(&nid_affected)
; return 0
; }
mclxUnary(cl_adj, fltxCopy, NULL)
; mclxMakeCharacteristic(cl_adj)
/* FIRST REMOVE ENTRIES set to zero (sssst now .. */
/* ...) and THEN make it characteristic again */
; mclvUnary(cid_affected, fltxGT, &bar_affected)
; mclvUnary(nid_affected, fltxGT, &bar_affected)
; *cl_adjustedpp = cl_adj
; *cid_affectedpp = cid_affected
; *nid_affectedpp = nid_affected
; return n_adjusted
; }
