►
A
A
So
there
is
a
written
version
of
this
update.
You
can
check
it
on
the
description
of
this
video
everything
that
you're
going
to
see
here,
you're
going
to
survive.
So
hopefully
so,
if
you
prefer
video
or
prefer
reading
over
watching
a
video
feel
free
to
go
over
there.
So
we
are
continue
working
exploring
envelopes
and
how
to
make
data
science
or
better
how
to
make
it
better
for
data
scientists,
because,
as
I
mentioned
before,
the
truth
is
data
scientists
don't
use
git.
A
Usually
when
you
finish
your
analysis
most,
you
would
have
to
write
multiple
reports,
because
multiple
people,
multiple
groups
of
people,
are
interested
in
the
results.
So
you
write
one
for
the
data
scientists,
so
a
more
technical
one
that
can
be
reviewed.
It
can
be
broadly
analyzed
the
code,
the
output
itself,
but
you
also
want
to
write
one
specifically
for
for
stakeholders
and
often
there's
more
than
one
type
of
stakeholder,
so
you
write
both
portfolios
and
what
happens
is
when
you
then
update
some
of
these
changes.
A
A
A
This
images
this
this
is
the
template
for
the
rhythmic,
and
I
add
these
two
different
tags
over
here
that
instruct
the
the
code
that
we
wrote
that
okay,
I
will
pick
up
an
image
from
this
jupiter
file.
That
is
key,
that
is
in
this
cell
and
with
this
type
and
now
I'll
insert
this
into
the
ring,
and
then
I
run
the
script
and
it
ends
up
with
this.
Is
the
jupyter
notebook
this?
If
you
look
at
the
at
the
raw
version
of
this
notebook,
you're
gonna
see
that
here
somewhere.
A
Let
me
just
look
for
this
here
yeah,
this
cell
has
a
cenoid
and
it
has
an
image.
So
this
is
what
I'll
pick
up
over
here
and
then
what
it
generates
is
this
rhythmic
file.
So
it
has
that
image
in
itself
and
the
table
itself.
So
that's
pretty
cool.
This
is
a
poc.
We
couldn't
be
generally
created
a
library
out
of
it.
A
We
had
this
tool
that
would
convert
their
notebook
into
a
markdown
that
would
make
it
a
little
bit
cleaner
to
to
the
diff,
but
well
now
it
did.
We.
We
use
this
to
create
the
depth,
not
only
that
implemented
in
a
way
that
it
can
be
used
directly,
so
we
use
git
drivers
and
this
can
be
used
either
as
a
dropping
replacement
for
in
itself.
So,
for
example,
let
me
open
over
here.
This
is
what
a
regular
div
would
like
look
like
without
a
our
changes.
A
So
you
see
a
lot
of
noise
here,
a
lot
of
different
things.
It
makes
it
really
really
hard
to
analyze.
What's
what
has
changed
now
you
see
with
what
we
did,
how
a
lot
easier
to
see
the
changes
that
happen
so
before
you
had.
Let's
just
see
how
many
lines
we
had
before
we
had
585
lines
in
the
previous
version
and
on
the
new
version
the
div
is
about
130
likes.
So
it's
20
it's
20
of
what
it
was
before.
A
A
This
is
red,
and
so
I'll
even
comments
by
github
folks
saying
that
they
want
to
see
this
on
github
as
well,
or
you
have
employees
so
we're
clearly
onto
something
here.
This
is
really
a
big
main
point
for
data
scientists
and
we're
on
the
right
way
right
now.
What
we
need
to
do
is
figure
out
how
to
integrate
this
with
hitler.
A
There
have
been
a
lot
of
conversations
with
leaping
and
other
teams,
so
code
review
team,
and
it
doesn't
seem
to
find
a
place
within
the
factory
that
you
have
right
now
or
there
is,
but
it's
not
optimal,
it
doesn't
really
matter.
The
point
is
how
to
debate
this
with
gitlab
is
a
problem
and
we
are
working
through
it.
So
we
expect
a
little
bit
lower
velocity
because
there's
a
lot
more
to
discuss
right
now
how
to
how
to
make
it.
A
It's
not
that
we
need
it's
clear,
that
our
users
need
this,
and
it's
clear
that
this
can
open
up
a
lot
of
more
doors
for
data
scientists
and
how
they
perceive
hitler,
and
that's
it
for
this
week
again.
Everything
that
you
that
you
watch
here
is
on
this
issue
and
we
have
the
tracker
with
the
issues
of
every
week.
So
if
you
are,
if
you
want
updates
on
what's
going
on
just
subscribe
to
this
one,
others.