►
From YouTube: SimPEG Meeting April 23rd
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).
A
B
Oh,
that's
a
lot
of
fun,
yeah
I
know
it
feels
like
a
long
time
ago
anyway,
yeah
well.
The
first
week
was
the
course
and
the
second
week
we
went
through
bueno
Cyrus,
so
it's
been
a
while
already
that's
am
it
was
overall,
pretty
good.
I
was
pretty
impressed
about
the
quality
of
students
in
Argentina,
whereas
they're
actually
having
a
hard
time
that
the
country
itself,
like
there's
a
big
inflation
going
on
so
I
mean
doing
science.
B
There
seems
pretty
pretty
hard,
but
that
so
they
are
actually
focusing
on
very
theoretical
works,
and
maybe
some
computer
related
works.
So
in
terms
of
that
their
background
about
physics
and
math
was
really
good,
which
was
yeah,
pretty
good
yeah.
So
the
course
went
well,
there's
three
other
courses
from
Mauricio
Saachi
in
Alberta
and
Nicholas
Lin
from
Switzerland
and
was
Juan
Alfonso
from
Australia,
so
two
of
them
were
deterministic
inversion
course
and
the
other
two
were
probabilistic
inversion.
B
The
course
materials
are
available,
so
yeah
I'll
make
it
available
through
the
course
website,
so
so
that
you
guys
can
take
a
look
because
it
was
actually
really
nice
to
kind
of
like
a
it's.
Like
a
time
stamp
of
what
is
the
state
of
art
and
the
tremendous
student
version,
and
also
like
a
probabilistic
inversion,
which
is
kind
of
emerging
I,
guess
in
these
days,
so
yeah
I
thought.
That's
that
that's
a
great
overview
on
my
end
anyway.
So
there
was
that's
great.
B
C
B
D
B
That's
the
main
thing
yeah
and
the
one
connection
that
I
want
to
make
the
example
that
I
mentioned
oh
I'll,
show
you
later,
but
that's
kind
of
there's
like
a
Nicklaus
meter,
an
example
or
they
find
like
kind
of
divided
to
region
and
then
so,
you're
you're
looking
for
a
boundary
and
also
the
distribution
of
each
boundary,
so
they're
inverting
like
using
MCMC
to
search
for
a
boundary
and
also
the
distribution
inside.
You
know
what
I
mean
cool.
B
Can
use
what
kind
of
dumbs
idea
we
can
combine
this
to
model
space
right,
but
you
can
have
one
model,
have
like
a
parametric
kind
of
two
region
and
is
working
for
a
conductivity
or
whatever
region
of
that
part
and
then
also
update
for
a
surface
and
then
got
another
model.
Just
updating
for
distribution,
for
instance,
so
yeah
I
mean.
That
seems
like
an
interesting
example.
So
he
mentioned
Oh
it'll
be
really
hard
to
do
using
the
term
in
second
version,
for
instance,
that's
what
he
sort
of
made
up,
but
then.
B
D
E
E
B
Yeah
so
I'm
sort
of
planning
to
so
this
that's
actually
was
something
like
what
I
am
moving
for:
I
kind
of
want
to
explode
of
probabilistic
approach,
MCMC
planning
to
use
pani
mg3
to
new
and
then
the
plugins
impact
as
a
black
box
code,
and
actually
it
requires
a
gradient
which
is
actually
kind
of
cool.
So
we're
ready
in
terms
of
that.
So
we'll
see
how
yeah,
how
far
I
can
go
about
that
I'm
planning
to
do
so.
F
D
Sure
so,
there's
two
releases
that
we
made
this
weekend
well:
I
guess
one
on
Friday
and
then
one
on
Sunday,
so
it
released
in
discs
or
ties
and
then
a
release
that
followed
in
simpe.
And
so
you
need
to
update
folks
like
simultaneously
or
else
you're
gonna
sorta
end
up
with
some
import
errors,
because
we
renamed
some
of
the
and
reorganized
some
of
the
base
classes
in
dis
critize.
It
shouldn't
change
any
of
your
scripts
or
any
of
your
notebooks
unless
you
were
doing
some
like
mesh
design
stuff.
D
So
if
you,
if
you
run
into
something
where
you
were
like
creating
your
own
meshes
and
there's
like
inheriting
the
base
classes,
he
might
friend
easier
problems
so
feel
free
to
ping
on
slacks.
If
you
do
but
I
think
most
people
are
just
using
what's
currently
in
discret
isin,
so
it
shouldn't
cause.
It
shouldn't
cause.
You
any
problems
there.
D
D
C
C
D
Fair
enough,
yes,
so
the
release
notes
are
out,
take
a
look,
some
of
the
oh
yeah
and
with
that
we're
trying
to
shift
all
of
the
documentation
over
to
the
numpy
styles
Docs,
rather
than
doing
like
the
base
or
the
standard
sphinx
stuff,
it's
easier
to
read
from
the
notebook,
and
although
it
like
in
the
code
I
think
it's
a
lot
easier
to
read
so
just
with
future
contributions
to
discretize,
try
and
keep
that
in
mind.
If
you
forget
about
it,
that's
totally
fine.
C
C
G
G
D
So
I
try
to
keep
the
pull
request
and
its
impact
pretty
light,
so
it
hopefully
shouldn't
conflict
with
any
of
your
branch
work.
If
it
does
just
ping
me,
and
we
can,
we
can
play
it
out,
but
it's
all
it
has
to
do
is
just
like.
We
put
the
bass
measures,
basically
under
discretize
dogface
dot
base
mesh.
D
H
Like
someone
had
a
question
about
saltwater
intrusion,
because
I
think
we
we
made
some
suggestion
that
we
would
try
to
study
it,
but
we
were
rather
ambiguous
and
truth
be
told
I
I'm
not
gonna,
be
sure.
How
is
the
tribe
characterizes
it's
all
right,
I,
don't
have
to
say,
like
you
know,
it's
something
we
have
to
think
about
a
bit
more.