►
From YouTube: SimPEG Meeting September 15th
Description
Weekly SimPEG meeting from September 15th, 2021
A
Hi,
hello,
everyone
I'm
at
ubc
today,
most
of
my
stuff
has
been
preparing
for
courses,
but
this
is
a
synthetic
meeting.
So,
let's
start
talking
about
some
civic,
see
add
your
names
to
the
attendance
list.
If
you
don't
mind,
add
quick
reports,
not
too
many
people
here
today,
so
probably
have
a
little
bit
from
everyone.
If
there's,
if
they
feel
like
sharing.
A
A
Doesn't
seem
like
it
okay,
so
I
will
start
with
some
quick
reports,
so
I
started
up
the
discretized
logo
doc.
It's
there's
a
link
there
in
the
notes.
A
A
Most
likely
just
agree
on
one
and
then
we'll
put
it
in
and
merge
in
the
documentation
updates,
which
should
happen
pretty
quick,
then
other
than
that
yeah
I've
been
a
lot
of
my
work
has
gone
into
preparing
for
the
class.
I've
been
able
to
devote
too
much
time
to
synthetic
at
this.
Second
there's
a
few
things
on
my
radar
as
far
as
a
few
pr's
that
need
to
get
a
last
minute
look
through.
A
The
only
other
thing
with
the
num,
with
the
numpy
documentation,
building
that
we've
got
with
all
those
stock
strings.
Devin
is
we
need
to
modify
the
build
so
it
it
is
building
correctly
on
the
on
the
testing
environment
right
because
we
were
it's
still
not
installing
that
collapsed
option.
So
it's
failing
because
of
that
on
all
the
everything
everywhere
it
tries
to
build
a
documentation.
That's
failing
because
of
that.
C
C
My
only
concern
with
the
some
of
this
is
to
make
sure
that
it's
still
a
stable
local,
build
on
any
operating
system.
You
know
so
yeah
we
needed
to
build
online,
but
it
would
be
nice
to
not
run
into
weird
finicky
things
if
someone's
doing
a
local
build
of
the
website
as
well.
A
The
time
so
at
least
that's
what
the
issues
that
you
were
having.
A
A
But
I
saw
when
I
was
putting
together
the
documentation
on
a
different,
a
different.
I
think
it
was
maybe
the
gpg,
not
the
gbg,
one
of
the
other
website,
documentation
that
I
was
working
on
recently
that
just
I
liked
how
they
did
it,
and
I
was
like
okay
well,
that
makes
sense.
A
C
Yeah
I
mean
I,
I
also
saw
some
other
options
for
collapsing.
This
was
just
the
easiest
thing
to
implement
if
it
already
exists.
Why
why
write
something
myself,
but
I
did
see
some
people
make
some
kind
of
extensions
or
whatever
it
is
they
they
made
something
in
the
in
the
repository
that
would
do
it
on
its
own,
so
they
managed
it
internally.
A
A
C
A
A
D
So
again
I
shared
that
on
slack
at
some
point
already,
but
it's
exciting
news
that
scikit-learn
have
decided
to
move
on
to
a
1.0
version
like,
and
I
think
lindsey
commented
also
that
it
seems
that
they
did
something
like
quite
free
planning
to
do
with
simping
and
dispatch
like
they
did
enough
that
they
can
consider
it
under
a
1.4.
D
I
started
going
through
that,
and
there
is
some
api
naming
convention
change
that
seems
to
break
pgi
at
the
moment.
So
I'm
going
to
work
on
that
and
it
should
not
be
too
bad
like
and
probably
I'm
gonna
include
that
in
the
pgi
improvement
branch
that
I've
been
working
on
lately
and
I'm
actually
gonna
focus
on
that
for
this
afternoon
and
and
the
rest
of
the
week,
whatever.
D
Whatever
this
api
change
things
and
add
what
I
wanted
to
add
to
that
pr
and
probably
get
it.
I
like
ready
for
review
by
the
end
of
the
week,
but
yeah
no
that's
cool
to
see
that
they
moved
to
1.0.
E
F
I
figured
I
got
that
mpi
problem
figured
yeah.
My
results
weren't
working
out
very
well,
because
one
of
my
nodes
was
yeah.
It's
just
not
it's
not
functioning
as
it
should
so
yeah.
The
npi
solution
is
working
pretty
well
now,
however,
still
it's
going
to
take
a
lot
of
restructuring
of
the
simpag
stuff
just
to
plug
it
in
that's.
Why
I'm
also
working
on
the
side
with
the
kubernetes
side
of
the
cluster,
because
that
it
just
plugs
in
really
nice,
but
right
now,
I'm
just
having
a
problem
with
setting
up
the
environment.
F
Correct
correctly,
I
just
can't
seem
to
get
the
right
sim
peg
to
load
up
everything.
I
was
actually
having
a
few
problems.
The
pgi
was
giving
me
some
issues
on
installing
it
on
like
a
pod
or
kubernetes
pod.
It
was
the
check
or
underscore
check
x
yeah.
It
was.
I
was
loading
in
the
update
or
like
a
really
recent
side
kit
and
that's
what.
D
Yes,
so
that's
one
of
the
change
that
they
made
in
1.0,
so
this
one
like
so
maybe
at
the
moment
a
quick
thing
is
like
I
think,
like
in
the
requirement
install
like
a
we
like
they
change
the
api
of
the
of
the
clustering
quite
a
bit
in
the
last
year.
D
D
F
Yeah
that
got
me
around
that
I
found
the
exact
version
that
you
had
on
up
on
the
requirements
so
that
solved
that
problem,
but
now
I'm
still
having
some
other
environmental
problems
yeah.
I
just
can't
can't
go
at
simpeg
to
quite
function.
F
I
can
launch
desk
like
just
a
simple
dash,
multiplications
and
simple
functions
like
that
on
the
kubernetes
cluster,
so
it's
just
getting
my
environment
ready
and
then
that
will
be
a
better
solution,
because
it
will
just
plug
right
into
the
code
that
don
and
I
have
there
on
child
simulations
so
but
yeah
I
got
up
to
that
and
then
I
started
going
over
the
code
cub
stuff.
Actually
I
might
bug
you
joe
just
to
see
if
I'm
covering
everything
right
and
understanding
it
later
today.
D
G
G
I
need
to
go
to
the
sparse
and
then
update
the
pgi
part
of
it,
but
it
should
be
a
lot
simpler
than
it
was
and
probably
gonna
need
another
weekend
to
be
able
to
get
through
it
and
that's
it
for
me.
How
do
you
add
the?
How
do
you
add
a
section
on
the
curve
notes?
Like
I
see
joe
and
tebow,
you
guys
have
like
a
block.
How
do
I
add
a
block
underneath.
A
There's
a
right
next
to
it
on
the
left
of
it
there's
a
you,
can
should
be
able
to
add
little
plus
buttons,
there's
little
plus
signs.
If
you
click
inside
a
block,
they
show.
C
C
Things
on
on
hold
one
thing
that
we're
working
on
and
that
we
may
end
up
putting
into
simpeg
at
some
point
is
working
with
sam
data,
so
you
would
have
a
a
time
domain,
em
source
and
you
would
you'd
have
a.
I
guess,
a
a
b
field.
Receiver
you'd
have
something
basically
like
a
magnetometer
and
it
would
be
measuring
the
projection
of
of
the
b
field
onto
the
earth's
field
and
then
yeah
working
working
with
gif
codes
to
be
able
to
model
that
and
interface
with
that.
G
A
Ben,
have
you
been
using
simple
much
recently
yep
yeah?
I
actually
found
a
couple
holes
in
the
desk
implementation
for
ip
we're
just
trying
to
incorporate
ip
into
our
drivers,
chairs
and
geo
apps,
so
yeah.
B
A
A
little
bit
of
work
on
our
local
branch
that
would
eventually,
hopefully
make
its
way
into
the
simplex
official
town
simulations,
but
yeah
cool,
so
just
by
holes.
Do
you
just
mean
like
areas
where
it
wasn't
doing
what
you
expected
it
to
or
yeah
that
it
wasn't
set
up
to
be
fully
tiled
for
ip?
I
guess,
yeah
cool
sounds
good,
looks
forward
to
seeing
it
when
it
eventually
gets
in,
hopefully
lindsay.
Have
you
been
able
to
do
anything
syntec
related.
B
I
gave
a
talk,
so
talked
about
it,
so
I
attended
the
aagc
on
monday
there's
a
workshop
on
airborne.
Em
I'll
share
the
slides
here.
If
anyone
wants
to
take
a
look,
it
was
a
really
good
workshop,
a
couple
of
the
things
that
were
well
so
the
sympath
or
talked
a
bit
about
sort
of
sim
peg
and
open
tools,
and
we
were
the
ones
who
went
first,
which
is
kind
of
kind
of
fun,
had
some
good
questions
and
and
good
discussion,
I
think
it
was
overall
well
received.
B
Jamin
was
asked
a
good
question
around.
You
know
some
of
the
terminology
about
joint
inversions
and
that's
something
tivo.
You
know.
We've
we've
talked
a
lot
about.
What
does
it
mean
to
be
doing
a
joint
inversion?
What
do
we
actually
mean
when
we
use
the
word
multiphysics
and
those
are
those
are
things
that
are
not
sort
of
well
established
in
in
the
community,
and
so
it
was.
B
It
was
good
that
he
surfaced
that,
and
so
we
had
a
bit
of
a
discussion
around
that
two
of
the
other
things
that
came
up
that
are
just
kind
of
one
thing.
That's
got
two
items.
I
guess
that's,
perhaps
just
something
maybe
interesting,
to
keep
on
folks
radars,
one
of
the
things
that
seems
to
be
quite
important
in
a
lot
of
work
in
australia.
B
C
E
No,
not
not
really.
I
I
just
follow
lindsay
around
and
check
her
talks
yeah,
but
there
is
some.
You
know
stuff.
That's
happening
a
lot
of
stuff
on
the
on
on
the
gif
tools:
side,
which
and
jiff's
side,
which
actually
has
some
major
implications
for
simpeg
development
and
funding
and
stuff
like
this.
So
that's
that's
kind
of
where
I've
been
investing.
My
efforts.
A
Well,
is
there
anything
anyone
else
would
like
to
share
any
thoughts
to
bring
up
I'll
post
a
link
to
that
google
doc
and
the
slack
there.
It's
already.