►
From YouTube: SimPEG meeting Sept 4
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
A
B
A
A
D
A
On
this,
so
this
is
the
the
current
like
live
version,
and
it's
not
it's
not
doing
exactly
what
we.
What
we
want
so
I
will
show
you
the
local
version
and
set
that
it's
on
there
a
tool
kit,
that's
your
site!
That
XYZ
is
it's
under
the
bigger
brother
Josiah
and
the
idea
is
to
give
simple
tools
so
that
geologists
can
start
like
when
used
your
physical
data
without
having
to
learn
like
a
whole
package
like
Joseph
or
arc
or
big
software
that
have
the
capabilities
to
do
most
of
this.
A
B
A
B
Comment
but
I
think
like
the
branding
for
the
bottom
saying
like
supported
by
MB.
Are
you
Jeff?
It's
not
perfect.
Okay
and.
B
A
Yeah,
absolutely
that's
great
and
there's
a
lot
I
think,
there's
more
or
less
given
to
the
side.
Look
all
the
titles
needs
to
be
short
time
and
don't
worry
about
the
site.
My
guess
is
a
statistic.
I
think
the
the
neat
thing
that
I
like
that.
Currently
we
don't
we're
not
supporting
on
the
light
version,
is
the
envy
sphinx
compiling
so
go
directly
from
a
notebook
to
rst
and
right
now
it's
doing
its
trick
in
terms
of
the
text.
A
Is
there
the
headers
are
there
from
a
notebook,
but
the
figures
are
not
or
not
showing,
and
this
is
I
think
has
to
do
with
any
strip
out,
because
I
give
the
same
problem
on
my
local
machine.
Oh
then,
you
need
to
strip
all
these.
We
need
to
go
strip
right
before
it
builds
up,
I'll
be
HTML
violence
and.
A
B
A
A
A
Try
to
do
that,
then
it's
going
into
the
packet
flow
yeah
we'll
have
to
figure
it
up.
Okay,
that's
a
good!
It's
minor,
minor
details,
logistics,
so
I
will
I
will
show
you
the
biblical
version
instead
of
the
the
library
on
my
team,
so
that
we
can
see
the
the
images
okay
and
then
these
are
technically
already
on
binder.
Absolutely.
A
C
A
A
Right,
so,
if
you're
just
your
Nero,
so
it's
like
this:
if
you
do
it
in
the
air,
it
looks
like
this.
You
back
those
the
first
three
are
using
Saipan
me
on
the
curvature.
I
am
coming
to
the
Riemann
curvature
and
right
now
it's
a
bit
slow,
but
it's
really
good.
You
know
K,
which
is
way
better
than
the
other
interpolators
and
there
mostly
to
deal
with
like
long
distances.
You
see
the
difference
between
cubic,
and
this
is
Cuban.
You
see
there's
a
lot
of
inline
kind
of
artifacts.
A
B
B
A
A
A
A
A
A
Bunch
of
averaging
applied
multiple
times
until
your
system.
Can
it
relaxes
to
there's
things
to
this
one?
This
one
is
it's
really
nice
when
the
points
are
really
really
scatter
mm-hmm,
because
it
takes
a
lot
less
time
to
your
affiliate.
Either
adds
action
takes
a
long
time
I'm
going
to
points
very
far
apart,
but
you
need
to
generate
the
advanced
system
and
then
that
eventually
gets
gets
too
big
for
that
gets
too
big
for.
A
A
2D
spline,
it's
a
2d,
spline
yeah,
so
you're
solving
for
a
quick
vigilance,
yeah
it's
like
a
distance,
analog
terminus
and
then,
once
you
have
your
prohibitions,
just
applying
it
I
think
goes
to
your
drilling
and
right
now
it's
pretty
slow
because
everything
I'm
doing
everything
in
four
groups,
but
I'm
sure
there
are
smart
ways
too:
yes,
okay!
So
anyway,
this
is
one
thing
that
we
should
you
thinking
about.
Bringing
on
is
their
critic.
C
A
B
C
B
And
there's
one
talk
in
the
kiyose
night
space
by
christian
thing,
she
said
Texas,
A&M,
I,
think
in
oceanography,
and
she
gave
a
really
good
talk
on
just
like
when
to
think
about
using
different
colored
wires
like
when,
should
you
be
using
diverging
versus
linear
I'm,
like
also,
she
had
a
really
good
discussion
in
there
on
like
when
to
use
nonlinear,
color
bars
and
like
when
it's
okay
to
highlight
these
things.
I
can
try
and
dig
up
that
link
to.
A
A
C
A
E
C
C
A
Everything,
okay,
I
was
here
so
okay,
so
basically,
Diane
was
interested
in
in
this
scenario:
we're
just
a
regular
positive,
magnet,
a
big
domain
and
some
demagnetization
inside
the
domain.
So
we
made
up
of
three
three
different
blocks:
basically,
one
like
this
one
big
that
is
like
way
bigger
than
draconian
and
one
that
is
trusted
in
like
this
and
people,
we
learned
on
Friday
that
y'all
just
interested
in
the
magnetization,
because
if
you
have
trust,
Falls,
yeah
fluids
and
that
that
breaks
down
magnetite
so.
A
A
A
A
A
A
A
What
the
app
role
that
you
do
is
carve
out
the
smaller
smaller
area,
so
you
can
do
that
when
go
around
and
right
now
it
will
break
if
you
have
holes
in
that,
so
I
mean
I
need
to
work
on
that
spit
down
to
get
around
them
and
have
2001
filters.
Okay.
Now
it's
this
on
the
viewing
against
reviewing
in
this
profile
up
against
and
then
just
poor
father.
They
that's
yeah
this
one.
Now
we're
adding
a
little
bit
more
functionality,
so
you
can
do
some
shading.
With
aside
what
orientation
does
on
you
wanna.
A
This
is
quite
nice
for
to
highlight
different
structures
and
in
Vinnitsa
chosen
depends
between
a
linear
of
an
air
color
scheme
versus
a
histogram
equalized,
and
this
we
should
add
to
the
synthetic
stack
to
we're.
Gonna
I
took
it
from
someone
else,
but
everything's
really
because
to
again
theta
in
linear
versus
equalize,
most
huge
difference.
A
A
A
G
A
A
B
A
A
Look
that
too
actually,
for
example,
this
is
a
horizontal
derivative.
I'm
at
and
upward
continues
a
little
bit.
Let's
see
the
content
will
change
right
so
now
we're
losing
the
high
frequency
quicker,
yeah
and
now
I
can
just
update
that
refresh
my
view
now.
I
get
that
so
people
can
start
playing
those
games
attenuating.
You
know
looking
at
deeper,
deeper
elements
by
move
in
here
Moving's
at
the
top
and
at
the
same
thing
here
you
can
can
highlight
the
time
things
from
you.
They
does
so
by
just
changing
this.
A
A
A
A
Yeah
this
guy
is
just
estimation
of
that
all
the
anomalies,
and
the
same
thing
is
that
let's
retain
the
color
and
further
they
denigrate.
So
we
can
see
the
contours
if
we
upward
continue
and
then
your
brush
now
we're
estimating
estimating,
like
just
larger,
larger
bodies
right
so
I
was
telling
as
that
in
this
region.
Yet
on
these
are
deeper
than
in
the
central
region,
where
it's
shallower,
volcanics,
so
Diane
has
been
took
this
and
on
the
side,
she's,
just
showing
people
how
to
that.
A
A
A
A
A
And
I
need
to
figure
a
way
to
be
able
to
do
the
FFT
holes
and
the
holes
in
our
drill.
I
think
it's
gonna
be
like
a
like.
An
interpolation,
probably
I
mean
on
curvature
inside
the
pattern
that
is,
there's
empty,
Judea
50
out
of
it
back
out
and
then
see
on.
Hopefully
there's
not
too
much
artifacts
or
on
the
around.
C
B
A
B
A
B
Yeah
and
then
I
think
like
as
much
code
as
we
can
sort
of
that
is
useful
to
bring
me
in
disinfect.
We
should
give
some
thought
to
that,
and
maybe
the
first
thing
to
do
is
and
just
like,
create
an
issue
on
simple
just
pointing
to
the
code,
and
we
can
figure
out
that
where
it
lives
oh
yeah,
because
then
we
at
least
like
just
document
that
this
stuff
needs
to
be
covert.
Mmhmm
yeah.
C
A
B
So
like
once
like
it's
sort
of
ready
for
prime
time,
we
should
think
about
that.
We
were
reading
everything
she's
very
played
a
leading
edge
style
article,
not
leading
edge
somewhere.
That's
geology
oriented,
but
I
bought
all
of
this,
because
this
is
I
think
that
would
be
a
good
way
to
spread.
The
word.
A
This
I
think
like
like
earlier
down
the
road.
If
something
else
you
know
picked
up,
yeah
come
on
aspects
of
this
project.
There's
like
in
branch
off
right
can
be
like
raggedy
processing
for
gravity
there,
and
you
learn
more
advanced
ones.
On
the
whole,
like
you,
we
were
talking
initially
to
do
the
education,
a
parametric,
inversion.
There's
like
way
more.
We
wanted
to
it's.
A
A
A
F
F
C
F
C
A
A
Thing
as
we
did
the
drape,
you
can
just
show
the
cross-sectional
inversion
as
an
overlay
on
Google
Earth,
which
would
be
a
good
way.
Don't
have
to
worry
about
like
3d
viewer,
Mensch
tools
or
whatnot.
We
just
export
section
and
say
I
want
to
see
that
section
I've
explored
up,
and
then
you
just
drape
it
thumb
on
the
top
right,
just
a
quick
way
to
expose
the
other
conversion
results.
A
B
A
A
F
F
B
A
C
B
B
B
B
A
B
And
I'll
give
you
some
pointers
for
creating
the
binder
links
and
then
also
I
think
it's
worth
creating
like
an
index
notebook
so
that
you
do
have
sort
of
a
showcase
for
when
people
first
hit
the
binder.
Instead
of
seeing
just
the
listing
of
notebooks,
they
actually
come
to
Lake
1804.
That's
got
some
description
and
pictures.
Okay,.
B
B
B
D
A
F
A
She
due
I
think
she's
gonna
be
do
like
at
the
end
of
October
or
something
like
I'm,
giving
herself
not
sure
yeah
which
restaurant
yeah,
so
it
would
be
I
think
she
probably
aiming
for
end
of
September
to
do
the
short
course,
and
that
would
give
us
two
weeks
to
just
really
wrap
it
up.
Whatever
makes
them
examination
they're
still.
D
A
A
C
A
A
Santa
there
this
is
when
you
look
at
VA,
and
we
also
we're
gonna
I'm,
also
going
to
add
profiles
and
they're
in
the
comments.
When
you
look
at
the
filters
like
it,
if
you
have
the
boy,
it
would
be
nice
to
see
it
right
because
yeah
like
right
sounds
a
little
gradient
like
okay.
What
does
that
mean
that
high
noise,
that
I
at
the
place
where
the
edges
are
or
it
starts
to
make
nice
for
us
bonnets,
I,
don't
know
yeah
I.
A
F
Yeah,
has
there
been
anything
done
on
that,
or
is
that
still
on
the
to-do
list
and
also
within
the
context
of
simply
what
actually
is
there
for
documentation
or
tutorials,
or
you
know
stuff,
with
respect
to
calculating
gradients
or
kind
of
working
on
up
on
a
mash
and
then
transferring
you
know
whatever
objective
function,
you
want
to
that
into
something.
So.
B
Here
we've
got
like
the
documentation
of
the
API,
but
that's
actually
basically
it
at
this
point.
There's
a
few
examples
that
show
different
regularization,
but
there's
not
much
actual
discussion
on
there's
not
actually
a
place
for
it.
So
this
is
one
of
the
things
I
think
we
really
need
to
think
about.
Is
this
all
go
on
to
discretize
so
every
well?
Not
all
of
it.
B
Doing
a
better
job,
organizing
the
docks
so
right
now,
there's
sort
of
it's
interspersed
like
information
about
each
thing
is
interspersed
with
the
API
and
I.
Don't
think
that
that's
how
we
should
be
doing
things
I
think
that
we
should
separate
it
out
like
I,
really
like,
for
example,
the
map
public
documentation
each
page.
B
Right
now,
for
example,
if
we
look
at
electromagnetics,
there
is
a
bit
of
theory
and
then
the
API
is
at
the
bottom
of
this
page.
So
it's
a
really
John
Ormus
page,
where
we
sort
of
have
Theory
interspersed
with
the
code,
and
so
here
we've
got
theory.
But
then
we
also
have
like
how
to
use
the
function
and
I
think
that
that's
not
how
we
want
to
be
doing
things
I
think
we
want
to
try
and
separate
out
theory
and
say
if
you
want
to
use
like
solving
a
problem.
B
B
If
you
really
need
that
level
of
detail,
we
don't
want
to
be
doing
that
in
this
MPEG
dogs,
but
we
should
have
usage
this
sort
of
just
talks
you
through,
like
here's,
some
things
to
think
about
it
in
a
magazine,
version
use
the
functions,
you're
gonna
need,
and
here's
maybe
a
small
example
and
then
have
the
API
separate.
So
if
people
want
to
go
in
and
look
at
actually
all
of
the
functionality
available,
they
can
I.
D
B
You
I'd,
so
that's,
oh,
that's
great
and
like
I,
think
sort
of
maybe
looking
at
this
there's
some
material
that
we
can
potentially
parse
out.
Yeah.
B
That's
that's
fine
and
there's
no
problem
being
overly
dense
because,
like
this
content
is
going
to
remain
useful
and
it's
just
a
question
of
where
it
lives
and
so
for
right
now.
This
is
this
is
perfect,
and
so
like
done
to
your
question
about
regularization,
like
first
of
all,
we
don't
even
have
a
top-level
title.
B
Which,
I
think
is
a
problem
it's
under
in
version
components,
and
then
we
do
have
regularization.
We
have
some
theory
discussing
taking
off
regularization
and
how
that's
discretize,
but
that's
that's
the
only
theory
that
we
have
for
the
regularization.
The
rest
is
then
documentation
of
the
API
in
which
there's
there's
some
information
interspersed,
but
again
it
is
kind
of
all
tangled
I.
F
Well
might
be
helpful
or
to
think
about
is
just
like
you
know,
a
series,
you
know
really
small
focused
notebooks,
like
okay
I've
got
a
specific
problem,
so
I'm
going
to
just
use,
let's
say
a
smallest
model
and
you
also
I've
got
you
know
a
variable
mesh
or
it
could
have
a
variable
mesh,
and
then
you
know
carrying
that
through
and
then
actually
showing
what
the
final
you
know,
discretize
for
where
that
is,
and
maybe
even
hooking
it
up
then
to
a
linear
inverse
problem
would
be
really
interesting.
C
F
F
That
you
are
in
counting
perhaps
come
back
to
just
the
definition
of
what
the
gradient
is
and
where
it's
living
right.
So
when
you've
got
adjacent
cells
that
are
really
different
in
size,
right
and
then
you've
got.
You
know
your
estimate
of
the
gradient,
but
you
know
you've
got
different
sizes,
so
there's
some
ambiguity
there
on
as
to
what
it
is,
you're
really
calculating
and
how
you
want
to
think
about
it.
F
So
that's
where
you
know
in
your
case,
you
were
effectively
getting
rid
of
that
change
in
scale
by
happy.
I
got
a
small
cell
here
in
a
big
cell
here
and
I
just
go
ahead
and
use
the
regular
gradient,
which
goes
for
the
middle
of
this
one
to
the
middle
of
this
one
and
then
put
that
in
you
know,
then
I
get
out
of
something
bait.
It
turns
out.
That's
perhaps
in
your
case,
that's
not
what
you,
what
what
you
want,
but
the
the
understanding
of
what's
going
on.
F
But
some
of
these,
at
you
know
a
fundamental
level
which
involves
you
know,
discretization
of
whatever
operators
you've
got
and
then
solving
either.
You
know
recomputing
the
integral
of
that
thing
or
using
it
in
an
inversion,
I
mean
those
are
really
kind
of
base
level
problems,
then
that
filter
up
into
everything
else,
so
I
think
those
are
kind
of
nice
to
have
in
mind
yeah.
B
B
Yeah
Geoscience
lecture
computation,
so
there's
a
question
of
like
what
this
is
actually
because
I
think
we
right
now.
We've
got
the
leading
edge
tutorial
in
there.
That
was
on
disk
because
in
the
DC
equations,
so
finite
volume
it's
in
there
we
have
a
little
bunch
of
stuff
like
sketched
out
and
so
I
think
yeah.
Let's
give
some
thought
to
like
how
this
interfaces
to
the
simpad
dogs.
Are
they
actually
the
same
thing
I
think
they
might
be,
and
how
do
we
go
about
and
merge
that
and
do
that
gonna
seem
like.
B
B
F
I
think
that's
really
hard
I,
that's
great
good
people
are
people
are
gonna,
be
excited
about
that
and
what
that
does
is
just
put
such
a
tangibility
on.
You
know
things
that
we
can
easily
talk
about,
but
nobody
has
really
got
a
sense
of
how
it
would
impact
I
mean
now.
It
can
in
fact,
that
you've
hooked
it
up
with
synthetic
and
with
the
search
data.
Do
various
things
overlay
that
on
Google,
Earth
and
that's
great
I
mean
well
I'm,
not
I
get
a
conversation
going,
but
actually
you
know
people
can
start
to
use
it.