►
From YouTube: SimPEG Meeting April 7th
Description
Weekly SimPEG Meeting from April 7th, 2021
A
A
B
Sure
yeah,
so
thanks
to
folks
who
were
able
to
join-
and
we
did
record
it
if
you
weren't,
I
linked
just
a
quick
update
or
sort
of
where
the
issue
for
tracking
the
governance
discussion
is
so
it's
in
the
community
repository.
B
So
that's
where
we'll
kind
of
formalize
these
sorts
of
things,
but
the
the
gist
of
the
conversation
was
that
we'll
sort
of
move
forward,
formalizing
two
of
the
roles
that
we
had
talked
about
in
the
governance
discussions
a
while
back
so
formalizing
the
role
of
a
managing
director
and
of
director
of
operations
so
joe
being
the
director
of
operations
myself
being
managing
director.
B
B
So
you
can
watch
for
that,
hopefully,
by
the
end
of
this
week.
That's
when
I
would
like
to
get
off
at
least
started
before
before
I
leave
on
friday
and
then
the
other
thing
that
came
up-
and
I
think
some
folks
have
added
some,
but
I
didn't
see
many,
and
so
we
might
postpone
this
for
another
week.
B
But
in
that
document
let
me
just
pull
that
up.
We
did
start
a
bit
of
brainstorming
around
a
mission
statement
for
simpeg,
so
we
captured
a
bunch
of
words
that
we
think
might
fit
in
a
mission
statement
and
want
to
sort
of
ask
everyone
if
you'd
be
willing
to
take
a
crack
at
writing
a
mission
statement.
It
doesn't
need
to
be
sort
of.
You
know
exactly
what
you're,
what
you're
happy
with
or
what
you
think
this.
B
This
should
be
in
the
final
form,
but
you
know
having
having
a
number
of
examples
and
sort
of
statements
to
look
at
and
compare
and
sort
of
see
what
what
fits
with
with
all
of
us
and
what
we
what
we
like,
and
what
we
want
to
highlight
would
be
great.
So
I
think
for
now
you
can
put
it.
I
think
that
the
best
place
would
be
in
the
document.
That's
linked
in
that
issue
that
I
started
and
I'll
just
drop
that
also
in
the
notes.
B
B
Yeah
I'll
drop
that
in
the
notes
and
in
the
chat
for
anyone
who's
interested,
I
see
tebow
tibo,
I
think,
has
taken
a
pass
at
one
and
so
I'd
invite
others
to
go
ahead
and
yeah
give
it
a
try,
see
what
you
think
and
we
can
sort
of
reconvene
on
that
in
the
next
in
the
next
few
weeks.
B
C
Maybe
what
you
discussed
about
having
a
big
day
one
year
three
year,
five
year,
plans
and
and
collects
certain
pr's
or
issues
to
to
get
into
the
next
version
kind
of
having
more
a
bit
of
structure
which
is
kind
of.
B
Yes,
exactly
so
that
was
kind
of
the
idea
of
both
planning
things
out
in
terms
of
kind
of
what
are
maybe
some
longer
term
goals,
and
also
thinking
a
bit
about.
How
can
we
better
communicate?
What
what's
coming
out
in
releases
like
what's
sort
of
on
the
collective
priority
list
of
everyone,
as
well
as-
and
this
is
something
that
would
be
great-
to
get
input
and
ideas
from
others
on
is
acknowledging
new
contributors
and
how
you'd
like
to
see
sort
of
new
contributions,
highlighted
celebrated
within
the
community.
B
We
talked
a
little
bit
about
maybe
making
use
of
the
simpag
twitter
account
and
having
sort
of
a
stream
show
up
on
our
on
our
website
of
of
that.
So,
if
you
have
any
other
ideas,
be
fun
to
get
creative
around.
You
know
celebrating
new
new
contributions
and
new
folks
joining
the
community.
So
definitely
give
that
some
thought
and
we
can
start
an
issue
to
track
that
as
well.
A
Let's
see
a
couple
other
agenda
items
at
some
point,
I'd
like
to
go
through
some
of
the
current
pr
statuses.
We
could
probably
do
that
after
the
meeting
today
when
we're
talking
about
coding
if
you'd
like
as
well
as
a
reminder
that
I
was
gonna,
I
said
last
week
that
I
have
a
open
coding
session
if
anyone
wants
to
stick
around
after
the
meeting
and
just
kind
of
hack
away
some
stuff
and
chat
interact
a
little
bit
yeah
there's
a
in
regards
to
the
pr
status
pr.
A
I
think
it
would
be
helpful
to
kind
of
just
go
through
some
of
them
and
have
an
update
about
where
they
are
and
where
people
what
people
need
to
do
with
them.
Still
there's
been
a
few
that
kind
of
got
stale
on
both
ends
as
far
as
reviewers
and
contributions.
So
it's
kind
of
a
general
reminder
of
try
to
stay
on
top
of
them
for
everyone,
including
myself,.
A
A
B
I
was
just
looking
for
my
the
link
to
my
slides,
but
I'm
giving
a
talk
tomorrow
with
folks
at
cu
boulder.
It's
the
computer
science
department
I'll
share
the
slides
around.
I
think
it's
open
to
anyone,
so
if
you'd
like
to
pop
in
you're,
more
than
welcome
to
yeah
and
I'll
share
the
slides
and
if
you've
got
any
suggestions
for
other
things
or
other
connections
you'd
like
to
see
made,
there's
some
good
groups
there
there,
where
there
are
a
fair
bit
of
overlap
or
sort
of
maybe
overlap
or
complementary
projects.
B
So
there's
the
earth
lab
at
cu,
boulder,
they're,
developing
a
lot
of
educational
resources
on
earth
analytics,
that's
led
by
leia
wasser,
also
pi,
open
sci
is
another
project
she's
leading
which
recently
got
some
funding,
and
so
pi
open
psi
in
a
sense
is
a
bit
well.
B
The
a
lot
of
the
idea
came
from
a
project
called
our
open
side
that
was
really
created
to
kind
of
create,
in
a
sense,
a
bit
of
a
mentorship
network
as
well
as
help
promote
sort
of
practices
and
standards
for
scientific,
open
source
software
and
promote
interoperability
between
packages,
as
well
as
to
grow
the
community
of
contributors
and
promote
more
diversity
in
who's
involved
in
scientific,
open
source
software.
B
So
that
is
also
being
led
at
a
cu
boulder,
so
very
excited
to
connect
with
them
as
well
as
there's
the
csdms
group.
I
think
that's
the
acronym
the
community,
it's
always
a
mouthful,
but
they
do
basically
modeling
and
sort
of
have
developed
some
software
to
try
and
create
like
a
layer
of
interoperability
between
surface
dynamic,
modeling,
and
so
that's
another
interesting
group.
B
I've
connected
with
them
a
bit
at
agu
before
so
yeah,
so
some
interesting
folks
to
to
be
connecting
with,
and
if
you
know
of
any
others
that
we
should
try
and
give
a
shout
out
to
in
the
presentation
that
would
be
that'd
be
much
appreciated.
A
And
then
you're
able
to
get
all
the
saudi
arabia
to
get
all
the
branches
switched
over
to
most
of
the
assuming
you
did
most,
if
not
all,
of
the
senpai
switching
the
masters
to
main
all
the
repositories.
B
Yes,
it
should
be
all
of
them
major
ones.
I
don't
know
if,
like
for
some
of
the
smaller
or
older
repositories,
we
I
might
have
missed
a
few,
but
all
of
the
major
ones
at
this
point
should
be
updated.
There's
a
line
thanks
for
bringing
this
up
joe
there's
a
line
that
I
put
in
the
issue
that
you
need
to
run
or
two
lines
that
you
need
to
run
to
get
those
changes
locally,
and
I
will
drop
that
into
the
media
notes
for.
B
D
A
A
Nice
devin:
do
you
mind
if
you.
E
A
E
I
just
thought
of
a
couple
of
things,
so
yeah
there's
the
em-1d
and
then
I
think,
a
few
months
ago
I
was
asked
to
put
some
work
in
on
the
self
potential
which
I
think
I
was
able
to
go
and
update
that
to
be
consistent
with
simulation
branch.
E
But
I
also
wanted
to
have-
I
guess
not
only
a
cell
centered,
but
also
a
nodal
formulation.
So
there's
still
some
work
left
to
do
on
it.
But
when
the
discretized
stuff
has
been
finished,
I
think
that
won't
be
too
hard
to
implement.
E
The
other
thing
was
some
conversations
have
gone
around
about
where,
where
we're
putting
data
for
tutorials
and
examples
and
things
because
we
don't
want
to
keep
large
data
files
in
our
repository
and
then
there
is
just
there's
lots
of
questions
about
you
know
accessing
and
where
to
put
it,
because
if
we
use
say
ubc's
cloud,
that
means
anyone
who
doesn't
work
at
ubc
can't
get
at
it.
So
lindsay
had
set
up
a,
I
guess:
a
google
storage
bucket
and
within
github
simpeg
community.
E
There
is
instructions
for
how
you
would
be
able
to
access
a
place
for
for
putting
putting
data
sets.
So
if
you
made
a
tutorial,
we've
kind
of
got
a
process
now,
where
you're
going
to
download
data
files
from
somewhere
else
and
unpack
them
and
then
run
the
script.
So
it
keeps
keeps
things
pretty
lean
within
sim
peg,
but
gives
us
a
place
to
put
to
put.
A
Things
that's
good,
and
the
other
point
is
like
we
would.
The
other
major
benefit
of
having
them
accessible
like.
That
is
that
if
someone
else
goes
and
downloads
the
script,
they
should
be
able
to
run
it
without
having
to
you
know,
run
it
in
a
specific
folder
inside
of
the
simpac
repository
or
let's
say
they,
you
know,
get
it
from
condo
they're
not
going
to
have
all
of
those
tutorial
data
sets
that
are
inside
the
repository.
That
would
be
expected,
so
it's
better
to
have
them
accessible
through
some
external.
A
Means
does
anyone
else,
have
any
reports
on
what
they've
been
up
to
some
type
related
things
or
just
kind
of.
D
D
As
far
as
I
know,
it
seems
they
were
going
for,
like
hybrid
exhibition
like
part
online,
but
in
person,
and
when
you
register,
you
can
actually
say
if
you
prefer
one
or
the
other,
if
you
or
either
one
so
actually
put
either
one
for
now
and
we'll
decide
later.
D
A
A
C
A
Quick
updates,
it's
okay!
If
not,
we
have
gone.
I
went
through
couple
of
the
older.
I
started
updating
a
couple
of
the
older
branches
that
I've
been
working
on
as
far
as
going
through
boundary
conditions.
I
put
a
few
more
updates
there.
As
far
as
testings,
I'm
I'm
still
having
a
test.
I
still
haven't
quite
figured
out
how
to
accurately
do
that
boundary
conditions
on
the
curvilinear
mesh
for
the
weird
conditions
like
it's
still
just.
A
A
So
it
should
actually
work
pretty
well
for
all
of
the
dc
related
stuff.
I
just
kind
of
did
some
quick
math
in
my
head,
so
he
I
did
get
a
chance
to
look
through
that
outline,
so
you
can
have
a
meeting
later
about
that
if
you'd
like,
but
I
I
have
a
clear,
I
have
a
clear
path
on
how
we
can
implement
those
other
boundary
conditions
for
the
nodal
problem.
A
F
A
Yep,
so
I
I
haven't
tested
it
yet.
I
just
basically
wrote
out
the
math
for
myself
to
see
like
oh
yeah.
This
is
what
I
have
to
do.
These
are
the
matrices
I
need
for
it.
I'm
like
okay,
I've
got
all
those
so
I'd
like
to.
Hopefully
we
can
my
my
thoughts
on
bringing
it
in
right
now
is
just
to
it's
close,
so
the
edge
like
the
edge
vector
integral
for
the
curvilinear
mesh
is
closed.
A
C
A
Well,
I
mean,
if
you
put
them
far
enough
away
like
it
still
hasn't,
it
always
has
an
effect
on
it.
I'm
going
to
make
it
like
it's,
but
if
you,
if
you
do
put
it
far
enough
away,
it
lessens
their
effect.
But
if
you
can
get
accurate
boundary
conditions,
you
can
really,
you
know,
tighten
up
the
mesh
and
you
don't
have
to
discretize
it
with
super
far
out
or
anything.
A
C
C
I
think
it
was
in
the
1d
case,
but
I
just
never
did
a
a
really
investigation.
So
I
wondered
if
someone
did
something
like
this,
how
does
the
result
change
if
you
swap
the
boundary
condition?
I
I
don't
know
would
be
interesting.
If
you
get
someone,
if
you
get
the
master
students
who
doesn't
know
what
to
present
the
next
scg,
get
him
get
him
to
use
all
your
boundary
conditions
and
play
around.
F
Yeah,
I
think
actually,
like
speaking
of
that
doing
that
for
mt
might
be
interesting.
I
don't
know
how
much
impact
you
would
get
for
control
source
because,
as
you
said,
if
you
got
a
big
enough
boundary
condition,
then
your
problem
is
solved
so
yeah.
A
A
Yeah,
I
do
remember
like
with
what
we've
done
with
the
mt
so
far
for
1d
and
2d,
the
boundary
conditions
worked
pretty
like
pretty
well
as
far
as
getting
accurate
solutions
for
the
layered
cases
and
kind
of
that's
good
moving
through
like
it
worked
pretty
well
so.
F
Speaking
of
that
joe,
where
are
we
on
that
2dmt
code?
Is
that
like
done?
Is
that
developed
or
more
things
to
do.
A
A
little
a
few
little
more
things
to
do.
Like
I
said
I
I
still
need
to
work.
I
still
need
to
get
the
receivers
hooked
up
properly
for
all
of
the
different.
You
know,
like
all.
A
A
A
A
B
Maybe
one
other
quick
thing:
this
is
also
a
question
for
soggy
is
transform,
is
coming
up
and
so
question
for
soggy
is.
Is
there
anything
you
need
from
from
the
group
here
as
you're
thinking
about
prepping
stuff
yeah?
I
I
think,
as
I.
F
I'm
just
I'm
keep
saying
I'm
starting.
I
first
started
I
loaded
up
the
rackmount
data
and
look
at
the
what's
been
done,
and
actually
I
got
a
quick
question
down
so
like
so
devin's
tutorial
is
great,
so
it's
actually
sort
of
shows
the
a
to
z
example
how
to
run
the
mac
code,
but
I
was
wondering
you
got
like
quite
a
few
drivers
and
kind
of
those
I
o
types
of
stuff
where
things
make
things
easier.
Like
okay,
you
got
a
data.
I.
F
E
Yes,
yes
yeah,
so
all
of
the
ios
that
I've
been
making
I've
put
in
simpeg
utils,
I
o
utils.
No,
no.
F
I
was
asking
more
like
an
example,
so
I
think
I
I
kind
of
walked
through
your
examples
and
it
looked
great,
but
the
like,
I
was
hoping
to
have
a
little
bit
simpler
version
like
okay,
like
just
generating
a
match
automatically.
For
instance,
I
think
adam
has
something
like
that.
I
guess
so.
I
was
hoping
like
to
have
some
sort
of
resources
like
that.
No.
E
E
So
I
I
didn't
make
any
like
simple
example
that
will
just
auto
generate
the
mesh
based
on
your
your
survey.
E
G
Published
publicly
before
the
sag
is
what
I
just
posted
in.
The
chat
is
basically
a
it's
a
script
right
and
then
you
have
it
an
input
file,
and
then
you
just
put
in
you
know
your
data,
your
mesh
specs,
and
then
you
basically
can
just
run
run
it
without
having
to
read
code.
F
I
think
at
this
point
like
so,
the
crowd
is
generally
geologists
or
scientists
have
like
maybe
that
much
geophysics
background
yeah
to
aim
more
conceptual
level
and
basically
running
the
code
and
changing
the
parameters
and
see
what
are
the
impacts
and
that
sort
of
my
goal
was
just
like
letting
them
know
that
kind
of
fundamental
ideas
of
the
inversion
like
data
missed
it
and
how
the
virtualization
works
and
sort
of
like
even
more
like
limitations
of
inversions
and
potential
and
possibility
by
like
sort
of
putting
a
relevant
information.
F
A
F
G
F
Yeah
yeah,
so
I'm
hoping
to
have
at
least
two
demo,
one
for
mac
and
one
for
em1d,
at
least
like
a
single
sounding
and
show
a
couple
of
inversions,
and
maybe
I
could
find
some
real
data
have
a
real
data
as
well.
So
I
can
show
120
comparison
or
something
like
that
and
doing
that
as
a
demo
will
be
nice.
G
Yeah,
so
it
looks
like
this
driver,
it's
pretty
old
right,
so
it's
it's
built
to
read
the
to
read
the
json.
Not
even
I
think
it's
just
a
text
file
at
this
point.
Okay,
yeah,
it
would
need
to
be.
G
You
need
to
be
reworked,
but
I
can
give
you
I
can
give
you
basically
a
a
python
file
that
reads
a
json
and
then
you
would
have
to
basically
just
change
the
change,
the
keywords
right
in
a
dictionary,
so
you
basically
have
a
dictionary,
just
change
the
keyword
and
then
rerun
rerun
the
function
would
that
would
that
work.
F
I
mean
at
this
point
like
the.
What
I
basically
need
is
just
a
mesh
generator
like
okay,
if
you
got
a
point,
the
magnetic
data.
Okay,
if
there's,
if
there's
an
automatic
mesh
generator,
I
mean
that's
probably
an
enough
for
me
to
start
with,
because
I
think
that's
most.
G
G
Yeah,
that's
good,
and
then
we
can
just
hide
the
credit
utils
and
just
hide
all
the
calls
on
under
the
basically,
you
would
be
supplying
points
and
then
the
type
of
the
type
of
parameters
for
each
and
just
fire
it
up
right,
get
the
mesh
back.
Yeah.
F
If
anybody
have
a
like
a
good
suggestion,
where
we
can
make
an
impact
to
either
geologists
or
whoever
job
geo
geoscientist
that
could
potentially
use
that
applied
geophysical
technique,
I'm
very
open,
because
I
don't
have
that
much
idea
at
this
point.
So
if
you
have
any
idea,
please
let
me
know:
that'll
be
that'll,
be
very
good.
G
Yeah,
you
need
3d
veils,
that's
the
that's
the
that's
the
point
of
connection
right.
You
want
to
see
it.
H
Do
you
have
any
ideas
about
like
who
will
be
attending
because
you
said
there's
going
to
be
geologists,
but
do
you
do
you
have
any?
You
know
further
breakdown
about
that?
I.
F
Know
I
asked
matt,
but
it
seems
like
it's
not
seems
not
very
clear
who,
like
it
seems
pretty
broad
and
yeah.
I
asked
matt
about
the
application,
but
yeah
it
wasn't
clear.
So
I
was
gonna
like
just
bit
broad
like
minerals
and
water,
not
sure
that's
really
hitting
the
target.
Peter
may.
C
Hard
to
say,
because
you
don't
really
have
to
register
you
can
watch
it
live
on
youtube.
So
I
don't
think
you
have
a
lot
of
hard
figures
except
for
the
ones
who
do
register
and
I
don't
think
you
have
to
sign
up
for
particular
tutorials.
You
can
just
join
the
channel,
so
I
would
expect
that,
like
in
the
few
days
before
the
tutorial,
you'll
probably
see
who
is
flocking
into
your
channel,
definitely
you're
on
tuesday
right.
C
I
don't
know
exactly
so
friday,
saturday.
The
hack
is
going
monday,
the
first
tutorial,
so
I
would
expect
that,
like
sunday
monday,
people
will
start
to
look
into
the
tutorials
and
join
the
channels
and
things
like
this,
but
it's
a
bit
late
for
you,
then
to
start
the
presentation,
then
so
yeah
it
doesn't
happen.
C
C
I
mean
lindsey
knows
probably
more
from
or
everyone
who
was
last
year
in
in
simpex
tutorial.
I
would
expect
a
similar
crowd
at
least,
and
you
were
there
too
soggy
right
so
yeah
I
mean
there
were
problems
from
how
to
install
sim
pack
to
more
advanced
inversion
questions.
You
have
the
whole
the
whole
spectrum
of
of
issues.
F
Yeah,
I
wouldn't
go
like
it
said
it's
two
hours
I
wouldn't
go
into
too
much
details,
probably
just
leave
that
very
high
level
and
just
like,
as
maybe
even
a
form
of
add
like
okay,
you
set
the
inversion
primary
run
and
see
how
that
looks
like.
Potentially,
we
can
show
some
sort
of
3d
view
as
well,
but
that's
probably
it.
C
Where
am
I,
and
anyway,
anyone
from
simpac
keep
the
time
free?
There
will
be
the
slack
channel
like
last
year
where
people
can
ask
questions,
but
then
there
will
also
be
gather
rooms,
so
people
that
have
problems
following
might
be
in
the
gather,
town
and
ask
per
video
chat
thing.
So
the
more
simpac
fox
is
around
the
better
to
help
stranded
lost
fox
to
get
installing
simpac,
and
things
like
this.
C
We
don't
really
know
how
that
turns
out,
so
maybe
no
one
is
there
in
gadra
and
everyone
in
slack
or
no
one
in
slack
and
everyone
in
gather
or
everyone
just
using
passively
on
youtube.
We
don't
know
we
will
see
next
year.
We
know
better,
but
you
should
try
it
out
next
week,
so
he
wants
to
stream.
If
you
have
time
one
day
next
time.
F
C
F
G
Well,
if
you
wanna,
if
you
wanna
start
right,
you
could
just
tweak
the
the
jupiter
app
that
I
have
on
geo
apps,
because
the
only
difference
for
you
is
you're,
probably
you
know,
you're
pulling
data
from
from
the
drive,
but
all
the
buttons
all
the
widgets
are
going
to
be
there
to
be
able
to
see
the
data
and
change
your
inversion
parameters
right,
you're,
probably
90
there
already.
F
If
there's
something
already
there,
I
might
just
use
it
that'll
be
great
I'll
I'll
point
it
to
you
and
yeah.
You
can
talk.
A
A
Are
there
any
what
kind
of
projects
are
involved
in
the
hackathon
so
far
or
people?
If
there's.
A
These
should
like
oh
yeah,
let's
get,
let's
try
to
set
up
a
group
or
serve
a
project
there
to
heck
away
at
something
on
that
day.
So.
A
E
C
I
can
here
are
the
the
hacks
that
are
already
proposed,
but
feel
free
to
propose
one
also
as
impact-specific.
A
Year,
like
this
I'll
copy
that
to
the
notes
as
well
just
to
have
them
there,
but
any
I
just
encourage
anyone
to
get
involved
with
that.
If
you
like
some
good
stuff,
I
think
so
from
one
of
the
ones
that
I
thought
I'm
interested
in.
Hopefully
I
can
get
at
least
help
on
the
discretized
side.
Is
the
two
sub
surface
from
subsurface
thing.
A
Also
part
of
that
I'd
like
to
support,
you
know
that
magic
tetrahedral
meshes
as
well
from
you
know,
from
like
gen,
pi
and
stuff.
Adding
that
kind
of
cross
compatibility
would
be
nice
yeah.
C
E
C
So
there
are
a
lot
of
tutorials
one
is,
and
it's
so
great
and
it's
actually
quite
heavy
on
em
and
potentials,
because
there
is
fatianda
there
is
pi
gimli,
so
we
are
probably
overrepresented
currently
at
this
transformer.
Please.
C
A
A
C
C
I
I
had
a
look
today
at
the
functions
of
mine
and
I
called
one
misfit
function,
but
it
could
have
equally
called
the
cost
function
or
objective
function.
Then
I
thought
about
and
optimize
I
think
in
simpek.
It's
also
called
optimize
right,
the
module,
but
then
we
usually
talk
about
inversion
and
then
yeah
I
mean
there
are
so
many
terms
floating
around,
depending
what
you
mean
and
where
you
are,
and
so
I
just
thought
that
I
I'm
wondering
what
people
think.
C
A
A
Okay,
say
big
goodbye
to
everyone
if
you
we're
gonna
stay
on
afterwards,
if
you'd
like
to
just
kind
of
chat
around
and
activate
a
few
little
things
here
and
there
for
senpai
related
stuff.