Changeset: 961473e2e4e4 for MonetDB
URL: http://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=961473e2e4e4
Modified Files:
clients/Tests/MAL-signatures_all.stable.out
clients/Tests/MAL-signatures_fits_geom.stable.out
clients/Tests/MAL-signatures_geom.stable.out
clients/Tests/MAL-signatures_none.stable.out
clients/Tests/exports.stable.out
gdk/gdk.h
gdk/gdk_batop.c
monetdb5/ChangeLog.Oct2014
monetdb5/modules/kernel/algebra.c
monetdb5/modules/kernel/algebra.h
monetdb5/modules/kernel/algebra.mal
monetdb5/optimizer/opt_prelude.c
monetdb5/optimizer/opt_prelude.h
Branch: Oct2014
Log Message:
Cleanup: removed BAThistogram.
diffs (truncated from 395 to 300 lines):
diff --git a/clients/Tests/MAL-signatures_all.stable.out
b/clients/Tests/MAL-signatures_all.stable.out
--- a/clients/Tests/MAL-signatures_all.stable.out
+++ b/clients/Tests/MAL-signatures_all.stable.out
@@ -232,10 +232,6 @@ comment Return the current size (in numb
command aggr.exist(b:bat[:oid,:any_2],h:any_1):bit
address ALGexist;
-command aggr.histogram(b:bat[:any_1,:any_2]):bat[:any_2,:int]
-address ALGhistogram;
-comment Produce a BAT containing the histogram over the tail values.
-
command aggr.jsonaggr(val:bat[:oid,:dbl]):str
address JSONgroupStr;
comment Aggregate the double values to array.
diff --git a/clients/Tests/MAL-signatures_fits_geom.stable.out
b/clients/Tests/MAL-signatures_fits_geom.stable.out
--- a/clients/Tests/MAL-signatures_fits_geom.stable.out
+++ b/clients/Tests/MAL-signatures_fits_geom.stable.out
@@ -232,9 +232,6 @@ comment Return the current size (in numb
command aggr.exist(b:bat[:oid,:any_2],h:any_1):bit
address ALGexist;
-command aggr.histogram(b:bat[:any_1,:any_2]):bat[:any_2,:int]
-address ALGhistogram;
-comment Produce a BAT containing the histogram over the tail values.
command aggr.jsonaggr(val:bat[:oid,:dbl]):str
address JSONgroupStr;
diff --git a/clients/Tests/MAL-signatures_geom.stable.out
b/clients/Tests/MAL-signatures_geom.stable.out
--- a/clients/Tests/MAL-signatures_geom.stable.out
+++ b/clients/Tests/MAL-signatures_geom.stable.out
@@ -232,9 +232,6 @@ comment Return the current size (in numb
command aggr.exist(b:bat[:oid,:any_2],h:any_1):bit
address ALGexist;
-command aggr.histogram(b:bat[:any_1,:any_2]):bat[:any_2,:int]
-address ALGhistogram;
-comment Produce a BAT containing the histogram over the tail values.
command aggr.jsonaggr(val:bat[:oid,:dbl]):str
address JSONgroupStr;
diff --git a/clients/Tests/MAL-signatures_none.stable.out
b/clients/Tests/MAL-signatures_none.stable.out
--- a/clients/Tests/MAL-signatures_none.stable.out
+++ b/clients/Tests/MAL-signatures_none.stable.out
@@ -231,9 +231,6 @@ comment Return the current size (in numb
command aggr.exist(b:bat[:oid,:any_2],h:any_1):bit
address ALGexist;
-command aggr.histogram(b:bat[:any_1,:any_2]):bat[:any_2,:int]
-address ALGhistogram;
-comment Produce a BAT containing the histogram over the tail values.
command aggr.jsonaggr(val:bat[:oid,:dbl]):str
address JSONgroupStr;
diff --git a/clients/Tests/exports.stable.out b/clients/Tests/exports.stable.out
--- a/clients/Tests/exports.stable.out
+++ b/clients/Tests/exports.stable.out
@@ -136,7 +136,6 @@ BAT *BATgroupvariance_population(BAT *b,
BAT *BATgroupvariance_sample(BAT *b, BAT *g, BAT *e, BAT *s, int tp, int
skip_nils, int abort_on_error);
BUN BATgrows(BAT *b);
BAT *BAThash(BAT *b, BUN masksize);
-BAT *BAThistogram(BAT *b);
BAT *BATimprints(BAT *b);
BAT *BATins(BAT *b, BAT *c, bit force);
BAT *BATintersectcand(BAT *a, BAT *b);
@@ -745,7 +744,6 @@ str ALGfetchoid(int *ret, int *bid, oid
str ALGfind(ptr ret, int *bid, ptr val);
str ALGfirstn(Client cntxt, MalBlkPtr mb, MalStkPtr stk, InstrPtr pci);
str ALGgroupby(int *res, int *gids, int *cnts);
-str ALGhistogram(int *result, int *bid);
str ALGhmarkp(int *result, int *bid, int *nr_parts, int *part_nr);
str ALGhsort(int *result, int *bid);
str ALGhsort_rev(int *result, int *bid);
@@ -2328,7 +2326,6 @@ int hasSameSignature(MalBlkPtr mb, Instr
int hasSideEffects(InstrPtr p, int strict);
str hashRef;
int headProp;
-str histogramRef;
int hlbProp;
int horiginProp;
int hubProp;
diff --git a/gdk/gdk.h b/gdk/gdk.h
--- a/gdk/gdk.h
+++ b/gdk/gdk.h
@@ -3190,17 +3190,6 @@ gdk_export int ALIGNsetH(BAT *b1, BAT *b
* @+ Common BAT Operations
* Much used, but not necessarily kernel-operations on BATs.
*
- * @- BAT aggregates
- * @multitable @columnfractions 0.08 0.7
- * @item BAT*
- * @tab
- * BAThistogram(BAT *b)
- * @item BAT*
- * @end multitable
- *
- * The routine BAThistogram produces a new BAT with a frequency
- * distribution of the tail of its operand.
- *
* For each BAT we maintain its dimensions as separately accessible
* properties. They can be used to improve query processing at higher
* levels.
@@ -3214,7 +3203,6 @@ gdk_export int ALIGNsetH(BAT *b1, BAT *b
gdk_export void PROPdestroy(PROPrec *p);
gdk_export PROPrec *BATgetprop(BAT *b, int idx);
gdk_export void BATsetprop(BAT *b, int idx, int type, void *v);
-gdk_export BAT *BAThistogram(BAT *b);
gdk_export int BATtopN(BAT *b, BUN topN); /* used in
monet5/src/modules/kernel/algebra.mx */
/*
diff --git a/gdk/gdk_batop.c b/gdk/gdk_batop.c
--- a/gdk/gdk_batop.c
+++ b/gdk/gdk_batop.c
@@ -1440,12 +1440,6 @@ BATmark(BAT *b, oid oid_base)
return bn;
}
-static void
-BATsetprop_wrd(BAT *b, int idx, wrd val)
-{
- BATsetprop(b, idx, TYPE_wrd, &val);
-}
-
#define mark_grp_init(BUNfnd) \
do { \
BUN w; \
@@ -1834,185 +1828,6 @@ BATpropagate(BAT *dst, BAT *src, int idx
/*
- * The BAThistogram function calculates the frequency distribution of
- * the tail values in its operand bat. Notice, that updates on the
- * result do not affect the delta administration.
- * Construction of a histogram over a string (or complex object) can
- * be sped up using the reference information in the BUN and bulk
- * copying the heap.
- *
- * There are separate versions for each type, and for each a hash- and
- * a merge-algorithms.
- */
-#define histoloop_sorted(BUNtail, cmpfnc, TYPE) \
- do { \
- const void *prev = BUNtail(bi, BUNfirst(b)); \
- \
- BATloop(b, p, q) { \
- v = BUNtail(bi, p); \
- if (cmpfnc(v, prev, TYPE) == 0) { \
- yy++; \
- } else { \
- bunfastins(bn, prev, &yy); \
- yy = 1; \
- } \
- prev = v; \
- } \
- bunfastins(bn, prev, &yy); \
- } while (0)
-
-#define histoloop_merge(BUNtail, HASHloop) \
- do { \
- BATiter bni = bat_iterator(bn); \
- \
- BATloop(b, p, q) { \
- v = BUNtail(bi, p); \
- if (BATprepareHash(bn)) \
- goto bunins_failed; \
- HASHloop(bni, bn->H->hash, r, v) \
- break; \
- if (r == BUN_NONE) { \
- /* not found */ \
- if (BUNins(bn, v, &yy, FALSE) == NULL) \
- goto bunins_failed; \
- r = BUNlast(bn) - 1; \
- } \
- (* (int *) BUNtloc(bni, r))++; \
- } \
- HASHdestroy(bn); \
- IMPSdestroy(bn); \
- } while (0)
-
-BAT *
-BAThistogram(BAT *b)
-{
- BAT *bn;
- BUN r;
- int yy = 0, tt = 0;
- BUN p, q;
- BATiter bi = bat_iterator(b);
- int tricky;
- const void *v;
-
- BATcheck(b, "BAThistogram");
-
- if (b->talign == 0) {
- b->talign = OIDnew(1);
- }
-
- if (b->tkey || BATcount(b) <= 1) {
- yy = 1;
- return BATconst(BATmirror(b), TYPE_int, &yy, TRANSIENT);
- }
-
- tricky = (b->ttype == TYPE_str && strElimDoubles(b->T->vheap));
- bn = BATnew(tricky ? (b->T->width == 1 ? TYPE_bte : (b->T->width == 2 ?
TYPE_sht : (b->T->width == 4 ? TYPE_int : TYPE_lng))) : b->ttype, TYPE_int,
200, TRANSIENT);
- if (bn == NULL)
- return bn;
-
- if (BATtordered(b) || BATtrevordered(b)) {
- /* the important information here is that equal values
- * are consecutive; we don't care about sortedness as
- * such */
- switch (ATOMstorage(bn->htype)) {
- case TYPE_bte:
- histoloop_sorted(BUNtloc, simple_CMP, bte);
- break;
- case TYPE_sht:
- histoloop_sorted(BUNtloc, simple_CMP, sht);
- break;
- case TYPE_int:
- case TYPE_flt:
- histoloop_sorted(BUNtloc, simple_CMP, int);
- break;
- case TYPE_lng:
- case TYPE_dbl:
- histoloop_sorted(BUNtloc, simple_CMP, lng);
- break;
- default:
- tt = bn->htype;
- if (bn->hvarsized)
- histoloop_sorted(BUNtvar, atom_CMP, tt);
- else
- histoloop_sorted(BUNtloc, atom_CMP, tt);
- break;
- }
- } else {
- switch (ATOMstorage(bn->htype)) {
- case TYPE_bte:
- histoloop_merge(BUNtloc, HASHloop_bte);
- break;
- case TYPE_sht:
- histoloop_merge(BUNtloc, HASHloop_sht);
- break;
- case TYPE_int:
- case TYPE_flt:
- histoloop_merge(BUNtloc, HASHloop_int);
- break;
- case TYPE_lng:
- case TYPE_dbl:
- histoloop_merge(BUNtloc, HASHloop_lng);
- break;
- default:
- if (bn->hvarsized)
- histoloop_merge(BUNtvar, HASHloopvar);
- else
- histoloop_merge(BUNtloc, HASHlooploc);
- break;
- }
- }
-
- /*
- * And now correct the interpretation of the values
- * encountered by bulk copying the heap as well
- */
- if (tricky) {
- assert(b->ttype == TYPE_str);
- bn->H->vheap = (Heap *) GDKzalloc(sizeof(Heap));
- if (bn->H->vheap == NULL)
- goto bunins_failed;
- bn->H->vheap->parentid = bn->batCacheid;
- bn->H->vheap->farmid = BBPselectfarm(TRANSIENT, b->ttype,
varheap);
- if (b->T->vheap->filename) {
- char *nme = BBP_physical(bn->batCacheid);
-
- bn->H->vheap->filename = GDKfilepath(-1, NULL, nme,
"hheap");
- if (bn->H->vheap->filename == NULL)
- goto bunins_failed;
- }
- if (HEAPcopy(bn->H->vheap, b->T->vheap) < 0)
- goto bunins_failed;
- bn->htype = b->ttype;
- bn->hvarsized = 1;
- bn->H->width = b->T->width;
- bn->H->shift = b->T->shift;
- }
-
- bn->hsorted = BATcount(bn) <= 1 || BATtordered(b);
- bn->hrevsorted = BATcount(bn) <= 1 || BATtrevordered(b);
- bn->tsorted = BATcount(bn) <= 1;
- bn->trevsorted = BATcount(bn) <= 1;
- bn->halign = NOID_AGGR(b->talign);
- if (BATcount(bn) == BATcount(b))
- ALIGNsetH(bn, BATmirror(b));
- BATkey(bn, TRUE);
- BATkey(BATmirror(bn), BATcount(bn) < 2);
- bn->H->nonil = b->T->nonil;
- if (b->ttype == TYPE_bit) {
- BATiter bni = bat_iterator(bn);
- bit trueval = TRUE;
- BUN p = BUNfnd(bn, &trueval);
-
- BATsetprop_wrd(b, GDK_AGGR_SIZE, (p != BUN_NONE) ? *(int *)
BUNtloc(bni, p) : 0);
- }
_______________________________________________
checkin-list mailing list
[email protected]
https://www.monetdb.org/mailman/listinfo/checkin-list
