<s>
In	O
statistics	O
,	O
biweight	B-General_Concept
midcorrelation	I-General_Concept
(	O
also	O
called	O
bicor	O
)	O
is	O
a	O
measure	O
of	O
similarity	O
between	O
samples	O
.	O
</s>
<s>
Here	O
we	O
find	O
the	O
biweight	B-General_Concept
midcorrelation	I-General_Concept
of	O
two	O
vectors	O
and	O
,	O
with	O
items	O
,	O
representing	O
each	O
item	O
in	O
the	O
vector	O
as	O
and	O
.	O
</s>
<s>
First	O
,	O
we	O
define	O
as	O
the	O
median	O
of	O
a	O
vector	O
and	O
as	O
the	O
median	B-General_Concept
absolute	I-General_Concept
deviation	I-General_Concept
(	O
MAD	O
)	O
,	O
then	O
define	O
and	O
as	O
,	O
</s>
<s>
Finally	O
,	O
we	O
define	O
biweight	B-General_Concept
midcorrelation	I-General_Concept
as	O
,	O
</s>
<s>
Biweight	B-General_Concept
midcorrelation	I-General_Concept
has	O
been	O
shown	O
to	O
be	O
more	O
robust	O
in	O
evaluating	O
similarity	O
in	O
gene	O
expression	O
networks	O
,	O
and	O
is	O
often	O
used	O
for	O
weighted	B-Algorithm
correlation	I-Algorithm
network	I-Algorithm
analysis	I-Algorithm
.	O
</s>
<s>
Also	O
implemented	O
in	O
the	O
Raku	B-Application
programming	I-Application
language	I-Application
as	O
the	O
function	O
bi_cor_coef	O
as	O
part	O
of	O
the	O
Statistics	O
module	O
.	O
</s>
