►
From YouTube: SimPEG Meeting August 18th
Description
Weekly SimPEG meeting from August 18th, 2021
A
Hello,
everyone
nice
to
see
you
all.
Let's,
let's
get
this
thing
going,
someone
just
added
something
to
the
agenda
just
to
count,
announcements.
A
A
C
Yeah,
just
do
a
quick
note
that
there's
the
scg
career
panel
happening
in
half
an
hour,
so
I
I
might
drop
off
for
that
and
if
anyone
else
is
interested,
it's
a
good
lineup
of
people
I'll
see.
If
I
can
dig
up
the
ad
and
I'll
drop
the
the
link
in
the
chat
here,
but
it's
more
mostly
the
gravity
I
think
electromagnetics
and
magnetics
section
or
folks
sort
of
in
in
those
spaces.
So
it
should
be
interesting.
A
Okay,
well,
then,
I'm
gonna
get
onto
some
quick
reports.
I
don't
really
have
anything
else
for
the
agenda
to
go
over
today.
I
kind
of
posted
it
a
little
bit
discretized,
but
over
the
unlike
the
discretized
channel
on
slack,
but
I
took
a
first
step.
A
I
was
already
doing
like
pde
solutions
with
it
should
slip
right
into
most
of
the
sim
bag,
structured
like
synthetic
simulations,
which
will
be
nice
motivation
just
to
be
able
to
support
triangular
meshes
in
general,
like
a
little
bit
more
conforming
to
topology
typography
things
enable
more
like
stricter
comparisons
with
other
the
other
tetrahedral
based
libraries
like
by
gimli
and
stuff
that'll.
Be
nice
yeah
it
it
came
together
quicker
than
I
thought
it
would
like.
It
was
pleasant
like
it
just
kind
of
like
started
working.
A
That
on
the
tetrahedral
meshes,
it's
basically
the
same
idea
that
also
led
me
to
realize
that
I
we
need
to
update
a
little
bit.
Those
two
like
face
inner
product.
I
train
our
product,
matrices
and
discretize
for
the
curvilinear
meshes
and
that
right
now
we
are
supporting
returning
like
an
inverse
matrix
for
them,
but
we
shouldn't
because,
like
they're,
not
diagonal
matrices,
so
we
can't
like
the
way
that
that
thing
returns
an
inverse
matrix.
A
A
The
face
description,
discrete
variables
on
like
em
anywhere
we
have
to
take
an
inverse
in
the
like
of
that
inner
product
matrix.
If
you're
looking
through
the
code,
you'll
see
it
it's
like.
A
A
A
A
That's
all
about
it
got
for
the
weekend
like
discretized
related
stuff.
It
looked,
I
don't
know
it.
I
was
happy
with
the
tetrahedron
that
should
be
able
to
get
them
pretty
quickly.
The
only
the
only
thing
on
it
that
I
don't
have
implemented
just
yet
are
is
like
the
interpolation
matrix
like
this,
that
arbitrary
interpolation
matrix
and
everything
else
kind
of
comes
together.
A
Okay:
okay,
lindsay:
what
do
you
got.
C
Just
a
quick
note,
I've
been
working
on
getting
the
line
current
source
up
and
running
for
the
hj
formulation,
so
when
current
goes
through
faces,
so
I'm
doing
that
right
now,
in
the
time
domain
and
I'll
take
a
look
at
the
frequency
domain
implementation,
it
should
be
straightforward
to
also
also
bring
it
in
over
there
and
then,
with
that
too,
I've
been
working
on
parsing
some
old
plotting
code
that
I
wrote
in
sort
of
my
own
personal
utility
library
for
fields
objects.
C
So
I'm
working
on
trying
to
like
parse
that
and
clean
it
up,
and
so
at
some
point,
I'm
hoping
to
generate
a
pull
request.
That'll
have
some
more
utilities
for
plotting
electromagnetic
fields,
I'd
sort
of
given
an
object
to
be
able
to
say
I've
got.
I
would
like
to
see
the
currents
or
the
charges
or
things
like
that
and
how
kind
of
reasonable
default
color
bars
and
default
choices
for
plotting
if
it
should
be
log
scale
or
not,
and
that
sort
of
thing
so.
D
C
Well,
that'd
be
awesome,
dev,
I'm
doing
a
pass
right
now,
so
I
first
need
to
I'll
get
him
into
a
digestible
form
and
then
and
then
we
can
touch
base
on
that.
I
appreciate
that
that'd
be
great
yeah.
A
C
Yeah,
I'm
I'm.
I
will
try
to
it's
a
fair
bit
of
logic,
so
it's
going
to
take
some
thought
to
like
parse
that
out
appropriately,
but
so
far
I've
at
least
got
it
ambivalent
to
if
it's
static.
So
if
it's
dc
frequency
domain
or
time
domain
that
that
can
all
work,
but
right
now
it
all
assumes
the
hj.
I
think
so
we'll
work
on
generalizing
that.
D
What
have
I
been
up
to?
Well,
I
guess
one
thing
while
lindsay's
still
here
is:
I
guess
I've
picked
up
where
liz
left
off
on
the
curve
note
project,
and
I
wanted
to
take
one
of
the
notebooks,
the
underlying
code
and
and
take
that
put
it
into
geoscilab.
So
I
just
get
to
call
the
app
in
a
really
clean
fashion.
D
So
I
made
a
pr
for
that
and
I
noticed
that
geoscilabs
had
not
been
updated
to
to
call
images
and
data
files
from
main
instead
of
master,
so
I
made
those
two
fixed,
the
underlying
codes
put
in
there
and
then
basically
updated
so
that
none
of
the
other
notebooks
break.
There's
a
pr.
That's
passing
that
I
I
provided
a
link
to
in
the
the
docs
for
this
meeting.
D
D
Oh
what
else
yeah
I
I
brought
in
your
your
changes,
joe
for
the
discretized
docs,
that's
great,
and
I
went
through
and
tried
to
find
any
little
spots
where
some
of
the
details
were
missed.
You
had
kind
of
a
itemized
list
of
things
we
wanted
to
make
for
the
documentation,
so
I
looked
through
that.
It
looks
really
good
and
I
also
played
around
with
using
the
collapse
this.
D
This
was
it
it's
sphinx
toolbox,
so
if
you've
got
really
long
plotting
scripts,
as
somebody
pointed
out,
I
think
last
week
or
really
long
pieces
of
code
that
really
don't
help
you
learning
that
function
they're
collapsed,
but
you
could
expand
them
if
you
want
to
so
I
like
it.
For
the
most
part,
it
looks
really
good,
but
that
toolbox,
I
think,
was
it's
compatible
with
sphynx
3.6.5
and
not
the
latest
version
of
sphinx,
and
I
wasn't
able
to
get
that
toolbox
from
like
installed
through
conda.
I
had
to
use
pip.
D
D
B
That's
I
think,
just
about
the
sphinx.
I
think
I
have
a
couple
of
things
extension
that
I
have
to
install
through
pip,
simply
because
they're
not
of
conda.
So
I
don't.
I
never
run
into
an
issue
if
just
an
extension
is
installed
through
okay
but
then
saying
same,
never,
never
yeah
yeah.
D
Yeah
I
mean
it
would
be
pretty
easy
to
remove
if
it
does
prove
to
be
a
problem,
and
I
did
find
some
stuff
where
people
had
just
written
their
own
little
extensions
to
to
just
do
that,
collapse,
yeah
and
then
also
just
looking
through,
and
I
think,
updating
the
the
dev
requirements.
D
There's
a
couple
things
that
we're
missing
so
yeah.
I
don't
know
if
what
you
want
to
do
is
a
final
review
joe,
but
I
think
we're
pretty
close
to
being
there.
D
Temporarily
exercising
the
the
tutorials
so
that
we
can
make
what
we
have
now
live
and
then
do
a
separate
pr
for
those
tutorials.
That
would
be
my
suggestion.
A
D
Yeah
I
mean
check
it
out.
I
didn't
go
crazy
with
it,
but
yeah
I
mean
there
was
some
things
with
highly
customized
plotting
that
was
taking
up.
You
know
a
whole
page
that
stuff
was
collapsed,
but
the
rest
of
it
was
left.
F
A
A
I
think
it's
going
to
look
good,
we'll
we'll
get
it
merged
and
then
probably
finalize
that
logo
and
then
tag
it
and
it'll
get
released.
D
A
So
we
can
just
add
a
little
bit
of
text
there,
just
like
oh
inheritance
diagram
for
something,
but
generally
they
are
at,
like
the
top
of
the
classes
like
when
you're
looking
at
classes
like
that's,
where
generally,
where
their
inheritance
diagrams
are,
but
usually
at
the
top
of
it,
and
then
goes
down
yeah.
So
yeah,
that's
fine!.
G
Hi
guys
yeah,
it's
been
a
while.
I
missed
a
couple
of
the
weekly
meetings.
Yeah
here
are
a
few
quick
updates
so
yeah.
Firstly,
I
just
I
removed
the
normalization
and
the
multiply
of
the
cell
volumes
as
the
drone
suggested
in
the
pr
and
another
thing
is
my
paper
got
published,
started
analyzing
the
uncertainties
using
the
mixed
lp
non-inversions
yeah
dominate
one
of
the
reviews
and
the
proposed
few
meaningful
suggestions,
yeah,
and
I
also
uploaded
the
code
used
in
the
code
and
the
data
used
in
this
manuscript
to
this
impact
research.
Community
yeah.
B
B
So
this
might
be
interesting
for
the
simpag
1d
stuff,
I'm
not
sure
in
the
long
run
it
just
well
the
ones
that
are
already
hardcoded.
You
don't
gain
anything,
but
if
there
is
a
new
filter
coming,
you
don't
have
a
good
field
to
someone
develops.
It
just
has
to
go
to
this
library
and
it
wouldn't
need
any
change
in
in
the
code.
So
that's
kind
of
neat
but
yeah.
I
just
thought
I
mentioned
it.
B
And
other
than
I
think,
because
soak
is
not
here,
I
started
to
pick
up
work
or
he
started
to
pick
up
work
again.
I
think,
partly
because
of
this
iaga
is
that
correct
lindsey
I
hear
poster
he
does
on
the
large
scale.
Yemen
version,
so
we
have
a
plan,
let's
see
how
fast
it
will
materialize,
but
there
should
be
a
plan
for
the
large-scale
yemeni
version.
B
And
before
I
forget,
because
it's
still
silent
the
book,
the
book
by
florian
wagner,
as
I
think
I
think,
that's
along
your
line
right,
tebow
things
like
you,
do
yeah.
E
Like
it's
it's
on
my
reading
list
for
the
for
this
week
for
sure
it's
gonna
be
interesting
too.
We
were
I
just
like
briefly
look
at
it
and
we
discussed
it.
We
are
it
with
lindsay
and
it's
it's
really
fun
like
it's.
It's
really
funny
that
we
really
have
like
like
a
lot
of
different
references
to
like
it's
very,
like
maybe
europe-centered
and
like
it's
me
like
so
it's
gonna
be
interesting.
E
E
Maybe
but
maybe
there
is
a
gap
to
to
bridge
here
between,
like
the
you,
like
the
effort
in
environment
and
the
one
in
mining,
maybe
maybe.
B
F
B
Multi-Method
inversions,
I'm
not
even
sure
which
one
they
use.
I
guess
dc
resistivity
along
those
lines.
I'm
yeah,
I
haven't
looked
closer
at
it
either.
G
B
C
For
publishing
a
book,
it
really
depends
where
you
go
through,
but
if
you
go
through
any
society,
they
usually
have
somebody
review
it.
It's
usually
like
a
bit
of
a
different
review.
I
mean
it's.
A
different
review
process
is
because
you
usually
end
up.
If
you're
going
to
go
through
a
society,
you
usually
like
pitch
the
book
to
them
and
it
sort
of
goes
through
a
pre
like
before.
You
actually
start
writing
it's
kind
of
approved,
and
then
the
review
process
is
much
more
to
like
improve
it.
C
B
C
One
other
thing
I'll,
maybe
just
put
on
folks
radars-
is
with
the
simpag
seminars.
If
there's
any
other
communities
or
people
you
think
we
should
reach
out
to.
Is
that
like?
Yes,
it's
sort
of
simple,
but
I
think
it
would
be
great
if
it's
also
simple,
plus,
plus
and
sort
of
reaching
out
and
making
connections
with
the
broader
ecosystem,
so
I've
reached
out
actually
to
santiago
who's
in
the
fariendo
community.
Because
there's
I
mean
there's
so
much
sort
of
like
commonality
between
the
problems
that
we're
solving
and
there's
such
complementary
packages.
C
It
would
be
great
if
we
have
yeah
if
we
can
use
this
as
a
bit
of
a
bridge.
So
if
there's
other
people,
you
have
in
mind
that
you
would
like
to
see
come,
give
a
talk
and-
and
maybe
try
and
strengthen,
connections
between
simpeg
and
and
elsewhere.
Keep
keep
that
in
mind
and
we
can
reach
out
and
yeah
and
try
and
use
this
as
a
bit
of
a
way
to
build
build
bridges
as
well.
C
Oh,
yes,
I've
posted
and
I'll
post
when
we
announce
xiaolong's
again
in
the
general
once
that's
ready
to
go
and
yeah
and
and
get
folks
connected
there
too.
So
yeah.
But
that's
a
great
point.
Tubo.
E
They
also
have
like
a
slack
channel
that
maybe
we
can
ask
them
if
we
we,
we
should
advertise
there
on
the
slack
as
well.
H
A
Okay,
yeah,
I
mean,
if
there's
anything
else,
floors
open
for
anyone.
Really
I've
got
otherwise.
I've
got
nothing
else
for
anyone
I'll
put
up
the
recording
again
right
away
this
week.
I
got
all
the
other
recordings.