|
||||||||||
PREV NEXT | FRAMES NO FRAMES |
getMaxNoClasses()
classes
Frame
owner.
Dialog
as its owner.
Frame
as its owner.
Frame
.
Frame
.
NPART
partitions in c
,
available in intersectMap
, according to the entropy
measure and distance measure
chrom2
based on
the fitness value of chromosome chrom1
and on the
intersection between chrom1
and the collection of
N
partitions in db
c1
relative to the partitions in the collection c2
;
N1
is the size of the collection c1
;
N2
represents the number of partitions in
c2
c1
based on the fitness value of chromosome
chrom1
and on the intersection between
chrom1
and the collection of N
partitions in db
v
with a Double representing the
value of the entropy of the partition determined by the
intersection of all NPART
partitions in the
collection c
and with an Integer representing the
number of classes of the intersection partition
c1
and classes of the partitions in
collection c2
p
and classes of the partitions in collection
c
silhouettes
of dimension 'size'
with the silhoutte coefficients, computed with respect to the
classes in this Partition and with respect to the values of the
NPART
partitions in collection c
; this
Partition should have at least 2 classes; neighbors
will contain for each rowid the class number of the closest
class
db
influences the target partition; the target
partition should be the last partition in db
;
sampleDBPct
specifies the percent of the database to
be sampled; sets wTargetCondonDB[i]
to
H(db[i]|db[target]) and sets the wDBCondOnTarget[i]
to H(db[target]|db[i]); sets weight[i]
to
wTargetCondOnDB[i]+wDBCondOnTarget[i]
fname
; the last column in db
is the
reference partition; NCOLS
is the total number of
columns, including also the reference partition
am
, with the
array representation of the partition; in this array each class
corresponds to an index in the array; the contents of the array
at that index is a Hashtable with the rowids where the class
appears in the partition
output
a graphical representation of the
silhouettes; classes are plotted in order and the elements in the
classes are plotted in decreasing order of their silhouette
output
output
output
nPart
partitions
sample
array, sampleSize
indices of the elements in values
that contain the
best values corresponding to an ascending order if
order
is 1 or to a descending order otherwise
n
integers between 0 and N-1
based on their probabilities; places the selected integers in
sample
; probabilities should be cumulative
probabilities
n
integers with uniform probability,
between 0 and N - 1
and places the selected values
in the field sample
rowid
to belong to class c
,
updates the class cardinality of the new and old class to whom
rowid belongs
am
that contains am.length
Hashtables
code
for the class
classOrig
c
to value
f
size
fName
into a collection of Partition objects
|
||||||||||
PREV NEXT | FRAMES NO FRAMES |