<s>
ArviZ	B-Application
(	O
)	O
is	O
a	O
Python	B-Language
package	O
for	O
exploratory	O
analysis	O
of	O
Bayesian	O
models	O
.	O
</s>
<s>
ArviZ	B-Application
offers	O
data	O
structures	O
for	O
manipulating	O
data	O
common	O
in	O
Bayesian	O
analysis	O
,	O
like	O
numerical	O
samples	O
from	O
the	O
posterior	O
,	O
prior	O
predictive	O
and	O
posterior	O
predictive	O
distributions	O
as	O
well	O
as	O
observed	O
data	O
.	O
</s>
<s>
The	O
ArviZ	B-Application
name	O
is	O
derived	O
from	O
reading	O
"	O
rvs	O
"	O
(	O
the	O
short	O
form	O
of	O
random	O
variates	O
)	O
as	O
a	O
word	O
instead	O
of	O
spelling	O
it	O
and	O
also	O
using	O
the	O
particle	O
"	O
viz	O
"	O
usually	O
used	O
to	O
abbreviate	O
visualization	O
.	O
</s>
<s>
ArviZ	B-Application
is	O
an	O
open	B-Application
source	I-Application
project	O
,	O
developed	O
by	O
the	O
community	O
and	O
is	O
an	O
affiliated	O
project	O
of	O
NumFOCUS	O
.	O
</s>
<s>
Numerical	O
summaries	O
and	O
diagnostics	O
for	O
Markov	B-General_Concept
chain	I-General_Concept
Monte	I-General_Concept
Carlo	I-General_Concept
methods	I-General_Concept
.	O
</s>
<s>
Integration	O
with	O
established	O
probabilistic	O
programming	O
languages	O
including	O
;	O
PyStan	O
(	O
the	O
Python	B-Language
interface	O
of	O
Stan	B-Application
)	O
,	O
PyMC	B-Application
,	O
Edward	O
Pyro	O
,	O
and	O
easily	O
integrated	O
with	O
novel	O
or	O
bespoke	O
Bayesian	O
analyses	O
.	O
</s>
