►
Description
Marcus Lewis shows off a tool much like the iPython Notebook, but for Comportex.
http://nupic2015spring.challengepost.com/submissions/37804-comportex-notebook
A
A
What's
going
on,
is
right
next
to
the
code,
and
I
think
that
this
is
a
good
way
of
bringing
them
together,
I
really
like
notebooks,
this
format
of
it
too,
like
in
case
anyone
doesn't
know
like
it's
sort
of
like
ipython
notebook
you
can
evaluate
is,
but
it's
in
the
document
form
other
concepts.
Just
real
quick,
comport
x
is
HTM,
implemented
and
closure
about
live
Felix
and
comport
X
vis.
A
Similarly,
he
made
a
bunch
of
really
cool
visualizations,
which
I
steal
profusely,
so
I'm
just
going
to
go
ahead
and
jump
in
so
I
just
want
to
get
something
on
the
screen
you
see
like
this
is
basically
come.
I
am
NOT
command
line
output,
but
you
see
this
model.
The
model
is
just
like.
This
is
a
spec
up
here:
it's
creating
a
model,
and
rather
than
evaluating
into
some
big
data
structure
or
variable
name,
you
can
see
it,
and
so
you
can.
You
can
like
toy
with
it.
A
A
This
was
just
like
a
toy
example
which
so
is
this,
but
I
decided
to
put
hot
gym
data
through
it
and
I
mean
this
was
all
today
starting
fresh
with
like
a
different
implementation
of
Mei,
new
comp
or
text
already,
but
starting
hot
gym
today
on
it,
I
didn't
I,
it
works,
but
I
tuned.
It
a
lot
more
if
I
had
a
bunch
more
time,
but
you
can
sort
of
see
the
investigative
process
of
how
you
would
start
tuning
it.
So
here's
the
model
I
started
with
I'll
just
go
on
run.
A
few
things.
A
Helper
functions
will
see
the
beginning
of
the
model.
All
of
these
are
red
because
it's
the
beginning,
they'll
start
to
turn
blue
and
purple
as
their
predicted,
and
so
it's
the
beginning,
there's
no
predictive
model.
Yet
another
thing
that
that's
going
to
be
useful
to
know
later
is
these
inputs.
The
way
encoders
are
often
like
this,
where
you
can
sort
of
read
them
like
reading
the
matrix.
This
one
is
the
time
the
hour
of
the
day.
A
This
one
is
whether
it's
the
weekend
or
not-
and
this
is
the
consumption
oh
and
I-
should
show
those
numbers
for
people
who
haven't
seen
hot
gem.
I'll
show
a
plot
of
so
the
corresponding
numbers.
To
that
were
the
red
line.
Is
that
actual
data?
Blue
dots
are
predictions
I
can't
predict
yet
because
there's
no
model
for
it
yeah,
it
hasn't
from
those
distal
synapses
I'll
jump
ahead.
A
So
the
reason
this
is
taking
time
is
because
closure
is
kind
of
cool
right
now,
there's
just
a
sequence:
that's
set
up
like
as
you
read
from
it,
it
evaluates
the
it
evaluates
the
simulation,
so
you
could
potentially
have
like
an
infinite
sequence,
which
represents
a
simulation
and
it
that's
just
a
fun
way
to
work
with
data
you
would
set.
So
this
is
500
time
steps
later
and
you
see
it
starting
to
predict
I'll
just
jump
head
to
a
thousand,
and
similarly
it's
about
to
do
500
type
stumps.
A
So
it's
not
a
it's,
not
that
good,
so
I
kind
of
like
to
drill
down
a
little
bit
and
see
what
the
columns
are
right
now
and
see
if
I
could
glean,
maybe
a
little
bit
of
insight,
so
what
it
would
be
a
good
time
yeah.
The
first
ten
should
be
good,
there's
plenty
of.
Let's
start
it
like
1004
and
do
like
sure,
10
steps
works
and
you
can
jump
up
and
down
and
we're
looking
at
the
early
ones.
One
thing
that
you're
seeing
here
you
there
are
different
visualizations.
A
A
So
the
reason
I
brought
up
the
encoders
is
when
you're
starting
to
look
at
columns
and
wondering
why
there
why?
Which
ones
are
predicted?
Are
it's
not
very
useful
to
look
at
ones
that
are
connected
to
the
upper
end
coders
only
and
because
it's
only
consumption
that
we
care
about?
That's
the
one
that
we
want
to
predict
correctly
and
so
I.
Let-
and
I
think
I
saw
that
towards
the
bottom.
It's
sort
of
a
it's,
it's
sort
of
topographical,
and
so
you
could.
A
You
know
to
look
at
those
and
you
start
like
looking
at
looking
around
seeing
which
ones
are
bursting,
the
red
ones
are
bursting,
and
so
you
see
this
one
was
somewhat
predicted.
I
think,
but
not
totally
so
you
start
getting
an
idea
of.
Is
this
being
over
predicted?
Am
I
seeing
a
lot
of
blue
blue
means
that
was
predicted,
but
not
active
red
means
active,
but
not
predicted,
and
it
just
gives
you
I
mean
you,
you
kind
of
learn
how
it
all
works.
A
lot
better,
just
through
seeing
this,
but
I
think
it
guides.
A
What
further
questions
you
ask
and
I
think
that's
like
a
really
useful
thing,
and
this
is
about
as
far
as
I
got
on
on
hot
gym,
so
I'll
just
go
ahead
and
leave
it
there,
but
I
think
you
can
see
how
that,
like,
I
just
said
how
it
would
guide
your
next
questions.
So
another
feature
we
will
I'm
going
to
and
save
this
save
your
file
in
c
oj.
A
So,
as
you
see
right
now,
it's
running
on
localhost,
which
means
it's
useless
to
most
the
world
right
now
and
I
will
drop
this
just
on
a
file
share
basically
I'm
to
a
static
website.
Okay,
that's
going,
and
so
what
I'm
doing
right
now
is
I'm
posting
this
to
the
web,
and
it's
going
to
be
in
the
different
form
where
it
sits
a
little
bit.
What
less
powerful
in
that
you
can't
execute
new
code,
but
has
it
done
yet
30
seconds?
A
So
basically,
what
I've
just
done
is
I
mean
I
was
writing
code,
but
I
was
also
writing
a
blog
post
or
an
essay,
or
you
could
slides
slides
for
a
presentation.
In
fact,
once
it's
there
and
10
seconds,
why
is
it
so?
Okay,
it's
a
little
bit
big
yeah
that
a
lot
of
the
work
was
making
it
were
these
weren't
enormous.
You
learn
a
hundred
ways
to
hang
a
browser
when
you're
dumping
new
pic
models
into
it
HTM
models
into
it,
so
it
should
be
there
now.
A
It's
downloading
that
oh
gosh,
hopefully
that
so
and
I
will
discuss
a
little
bit
more
about
like
I,
said
it's
running
on
the
the
HTM
in
this
case
is
running
on
a
server
and
visualizations
and
the
browser
which
is
new
for
its.
That
is
basically
the
model
you
would
use
with
new
pic
with
comport,
X
and
visualizations
of
it,
because
it
conveniently
can
compile
down
to
closure
script.
A
All
the
simulations
up
to
this
point
ran
in
the
browser
and
which
is
like
often
good,
but
they're
often
reasons
you
would
want
to
do
it
this
way
as
well,
and
that
was
a
big
part
of
the
work
here
was
making
that
possible.
So
it
loaded
as
you
can
see,
it's
no
longer
like
this
text
in
foot
area,
but
I'll
go
way
down
to
this
cool
one.
A
This
one
we
were
just
yet,
and
it's
still
interactive,
so
you
can
post
a
link
to
these
and
like
it,
you
just
saw
me
create
it
in
front
of
you
and
I
think
it's
just
opens
up
new
scenarios.
It's
good
to
get
away
from
localhost
so
and,
like
I
said
that
was
that
was
the
trivia
a
little
bit
different
from
comfort
xmas.
So,
as
mostly
all
I've
got
I
think
seeing
is
often
really
good
at
making
you
notice
questions
you
should
have
been
asking
or
that
you
didn't
know
to
ask
and
notebooks
are
really
good.
A
Dev
environment,
like
what
you
just
saw,
was
a
dev
environment
that
was
like
I
wrote
code
in
that
constantly
this
this
weekend
and
what
these
powers
combined
it
one
way
that
this
is
I
would
say
more
convenient
than
comport
X
vis
is
in.
We
we
were
both
struck
by
this
is
like
we're
confused
by
something
those
going
on
there
and
that
we're
like
wait.
We're
in
a
coating
place
right
now.
We
can
probe
in
right
now
and
like
look
at
it.
A
B
A
Here's
not
good
call
yeah,
it's
a
comport.
Sorry,
it's
a
gorilla
repple,
which
is
basically
someone
saw
ipython.
Notebook
he'll,
probably
watch
this
at
some
point,
because
I'll
use
this
to
make
them
accept
my
pull
requests
but,
but
anyway
and
he'll
see
that
this
is
getting
recursive
so
but
yeah,
it's
called
guerilla
repple.
It's
the
similar
one
for
closure.