►
From YouTube: SimPEG meeting March 27
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
C
Yeah
so
I
was
his
PhD
student
in
geophysics
for
a
bit
so
I
up,
geophysics
background,
but
now
I
do
mostly
coding
at
sequent
working
on
the
Python
side
of
things
and
my
title
is
integration
lead
so
I
do
like
sort
of
partner
integration
stuff
and
like
API
clients
and
getting
people
external
the
company
using
our
stuff
and
I.
Also
I
guess:
I
wrote
properties
which
a
lot
of
synthetic
is
now
kind
of
built
on.
So
that's
my
biggest
sin,
big
connection
that
satisfactory.
B
Actually
well
so
the
way
we
structured
these
meetings
for
those
of
you
who
are
new
we'd
start
with
just
like
quick
updates,
so
we
get
sort
of
a
60
second
max
update
from
people
who
have
stuff
that
they
want
to
talk
about
and
then
we'll
move
into
the
longer
discussion.
So
I
can
maybe
give
the
first
point
and
feel
free
to
jump
in
here
too,
because
you're
involved
in
this.
B
This
week
we
started
a
new
github
organization
called
GSI
data,
I
think
it's
Josiah,
GSI
xyc
data
I
need
to
double
check,
I,
think
it's
GSI,
xyzt
beta,
and
so
this
can
be
a
in
github
organization
where
we
can
post
data.
So,
for
example,
right
now,
Mount
Aiza
the
data
are
there
in
a
single
repository.
B
B
Yes,
so
it's
meant
to
be
a
place
where
people
can
put
small
ish
datasets,
so
it
shouldn't
be
enormous,
but
something
that
can
easily
be
hosted
on
github
is
welcome
to
go
there
I
think
right
now,
only
DOM
and
I
have
admin
access,
so
just
ping
either
of
us.
We
can
add
anybody
at
some
point
this
weekend
or
something
like
that.
We
can
try
and
add
all
the
simpang
group,
but
if
you
need
something
more
immediately,
just
paint
one
of
us
I.
D
B
Because
individually,
so
like
we
don't
we
don't
own
those
data,
so
we
need
to
honor
whatever
licensing
restrictions,
miss
Peregrine
or
whoever
owns
them
wants
on
them.
So
that'll
give
us
like.
If
each
data
set
is
under
its
own
control,
then
licensing
is
much
more
clear,
so
we
can
maybe
move
like
either
treat
the
GSI
slash
data
as
just
like
a
list,
or
it
could
eventually
be
the
like
website
that
points
to
all
the
different
data
sets
instead
of
in
the
GSI
data
org,
is
that
yeah.
B
D
That's
pretty
cool
because
we
can
use
the
interaction
to
be
able
to
do
the
slicer
at
the
same
time
as
using
the
the
ipython
widgets,
so
both
are
connected
at
the
same
time.
That
was
pretty
cool,
so
we
can
do
basically
for
its
simulation
of
the
of
the
block
at
the
same
time
as
as
the
lake
there.
So
we
change
the
change
the
model
and
see
the
magnetic
response
at
the
same
time,
and
then
from
there
we'll
just
be
able
to
do
Runyon
versions
and
then
show
truly
result.
That's
gonna
be
the
idea.
D
D
Yeah
well
also
congrats
to
you
to
John
John
I,
finally
managed
to
get
the
get
the
girl
running
so
so
now
we
have
very,
very
geometry:
I
I'm
not
going
to
show
you
anything,
it's
the
same
thing
as
as
last
week,
but
I
started
a
pool
request.
Another
question
is
there:
is
that
the
whole
bass
bass
thing
that
I
just
already
started?
Did
you
want
to
do
two
PRS
or
because
it's
gonna
be
a
lot
of
work?
D
D
A
D
E
E
Rather
short,
yeah
I
got
some
good
contacts
afterwards,
there's
some
people
who
had
some
recommendations
for
borehole
instruments
and
I.
Think
given
our
budget
in
scope-
it's
probably
not
so
applicable,
but
it's
nice
to
have
contacts
and
I
also
did
have
the
opportunity
to
talk
for
a
good
time
with
a
program
officer
for
what
to
do.
Scientists.
A
F
G
G
So
I
was
actually
using
octree
code
and
I
use
using
dumb
smash,
elder
I
have
to
change
a
couple
of
things,
but
this
is
the
power
line
that
I
have
built
up.
So
it's
actually
pretty
small
mesh
like
a
couple:
centimeter
mesh
or
the
par
line
discritization,
but
the
full
mesh
is
pretty
large,
so
I
I
think
I
ended
up
using
about
a
million
cell,
but
I
can
actually
do
this,
which
I
thought
pretty
cool
and
like
at
at
the
end
of
the
day.
G
What
you
get
is
something
like
this,
so
this
is
the
simulated
data
and
that's
a
filled
data.
So
well,
it's
not
exactly
the
same,
but
it's
pretty
similar,
meaning
like
a
you
can
actually
model
the
pawn
line
effects
from
from
the
India
airborne
data,
which
is
I,
thought
pretty
pretty
cool
and
yeah
without
it.
Without
the
re
code,
it'll
it'll
be
much
hard.
It
would
have
been
much
hard.
So
yeah
I
really
appreciate
all
people
when
developers
heard
me
what.
G
G
H
B
So
just
a
quick
update
from
you
for
me,
I,
won't
necessarily
show
this
right
now
like
following
up
from
last
week's
ondrea's
question
about
running
anisotropic,
I
went
through
and
just
created
a
simple
example
that
compares
to
a
domain
simulation
in
3d
with
simpe
within
my
Mont
simulation
that
runs
one
year
and
we're
all
really
very
slow,
which
is
exciting
and
the
right
now.
It
was
a
very
coarse
distribution.
So
it
runs
a
bit
quickly,
but
it's
encouraging
to
see
that
yeah
they're,
producing
similar
results
so
and
I
start
looking
and
feel
free
to.
B
H
H
That's
kind
of
yeah
I
wanted
to
maybe
just
give
like
a
quick
overview
of
some
of
the
work
that
we're
doing
at
sequence,
with
the
open
mining
formats
with
a
global
mining
standards
group
that
might
be
of
interest
to
some
people
here:
who's
who's,
actually
heard
of
OMF
or
global
mining
standards.
Group
at
all.
A
H
So
this,
the
global
mining
standards
group
is
sort
of
industry
standardization
body
for
dealing
with
mining
operations,
so
they
do
everything
from
sort
of
the
exploration
side
right
down
to
sort
of
operations.
So
that's
like
streaming
data
from
like
people
and
machines
going
around
mine
sites,
and
so
what
we're
we've
been
involved
with
sort
of
sequence
and
leapfrog,
which
is
a
geologic
modeling.
Software
is
on
the
data
exchange
for
mine,
mine
data
and
sort
of
that
covers
exploration
to
everything
through
to
production,
but
like
less
so
on
the
operations
space.
H
So
streaming
data
isn't
really
in
in
this
piece.
So
this
be.
The
idea
here
is
to
like
build
sort
of
this
open
source
file,
interchange
formats.
So
a
lot
of
the
software
packages
in
the
industry
don't
communicate
well
in
terms
of
linking
up
workflows
between
two
sort
of
proprietary
formats,
and
so
that's
something
where
the
GMG
is
trying
to
step
in
propose
an
open
mining
format
so
that
we
can
actually
do
that
franklin
is
there
anything
you
want
to
add
to
that.
C
H
No
I
think
it's
it's
like
totally
cool
to
see
like
these
very
heavily
proprietary
software
companies
and
mining
companies,
sort
of
both
Franklin
and
I
have
been
hammering
through
on
the
open
side
of
it,
so
that
we
can
actually
get
to
open
standards,
as
well
as
open
developments
of
up
this
file
format
so
like
getting
all
of
it
onto
github
having
contributor
Guidelines
code
of
contacts,
all
of
all
of
that
sort
of
stuff
that
you'd
expect
from
an
open
source
project.
So
so
that's
that's,
been
pretty
cool.
H
Know,
there's
probably
like
a
ton
of
different
like
open
standards
which
makes
them
not
so
standard,
but
this
this
one's
specifically
focused
around
the
mining
data,
which
has
some
more
specific
needs,
especially
around
sort
of
storing
big
hold
on
block
models
which
are
like
equivalent
to
our
tensor
meshes
or
regular
meshes.
So
that's,
that's
sort
of
where
they're
coming
at
it,
but
we're
also
trying
to
work
with
other
standards
bodies
like
the
open,
geospatial
consortium
in
terms
of
just
some
level
of
interoperability.
Even
if
it's
at
like
naming
conventions
at
the
stage.
Okay,.
C
H
G
H
F
G
H
So
this
this
is
like
point
points
that
data
line
set
data
which
can
be
sort
of
the
GIS
lines,
but
I'll
show
you
maybe
a
different
version
of
that
in
a
little
bit
and
then
the
surface
data,
which
is
just
the
triangulated
ones,
and
so
the
line
set
one
is
what
is
currently
being
used
for
things
like
drill
holes
and
all
it
is.
Is
this
sort
of
vertices
and
segments
definition?
H
G
G
It
just
got
a
quick
question:
okay,
even
just
step
back
so
they're
like
a
tons
of
formats
right
so
even
before,
let's
say
the
well
data,
there
are
a
lot
of
like
formats
and,
what's
your
vision
to
handle
like
it,
somehow,
you
need
to
transfer
that
into
this
format
right.
What's
up,
what
are
the
vision
here?
How
how
are
you
seeing
this
problem
to
work
it
out
so.
H
I
guess
a
lot
of
those
again
this
like,
whereas
this
is
sort
of
being
position,
is
on
a
data
transfer
between
primarily
the
mining
software
packages
in
the
industry
today,
and
so
that's
that's
the
the
problem
that
it's
trying
to
solve.
I
think
where
it
gets
a
little
bit
more
interesting
is
on
the
the
volume
like
block
model
storage
so
and
then
in
these
other
standards.
H
In
terms
of
like
last
data
or
some
other
drill
hole
database,
we
would
eventually
go
through
and
add
that
sort
of
extra
layer
that
hopefully
coincides
with
or
is
interoperable
with
those
standards,
okay,
I'm.
So
it's
so
it's
it's,
allowing
the
mining
industry
its
own
space
to
play
a
little
bit
and
then
sort
of
standardizing
over
time.
That's
a
bit
of
the
vision
weight
there.
There
really
isn't,
especially
on
the
sub
block
model,
anything
out
there
and
that's
the
primary
pain
that
the
mining
industry
is
feeling
at
the
moment.
Yeah.
B
We
might
consider,
when
sort
of
rewriting
the
data
class
or
like
what
that
should
look
like,
because
that,
in
some
sense
this
is
like
a
point
set.
Maybe
it's
a
more
complicated
point
set,
because
it's
like
a
time
series
of
each
point,
but
are
there?
Is
that
sort
of
in
scope
and
then
also,
even
if
it
isn't?
What
kind
of
things
do
you
think
like?
We
should
have
in
mind
in
developing
that.
H
So
it's
it's
not
really
like,
specifically
in
scope
its.
It
is
like
in
the
more
general
sense
in
the
in
that
you
can
have
points
with
arbitrary
metadata
associated
with
them,
so
it
could
be
a
output
format
for
simpang
and,
like
the
like,
the
container
that
it's
writing
to
is
not
going
to
be
overly
glorified.
So
it
could
be
better
to
maybe
right
to
hdf5
or
something
like
that
for
from
the
same
peg
side.
H
But
having
that
standardized
thought
about
how
like
the
primary
drivers
between
the
design
of
this
or
Franklin
and
myself
and
so
I
think,
like
we've,
we
have
put
a
lot
of
thought
into
how
to
sort
of
like
bind
data
to
these
things
in
the
metadata
space
as
well.
So
there
could
be
like
ideas
that
you
pull
from
there
make
sense
yeah.
B
H
C
Sure
so
yeah
so
I
mean
the
view
lfyou.
The
thing
that
we
made
that
I'm
gonna
be
talking
about
is
just
basically
like
a
online
viewer
for
3d
data,
where
you
can
upload
your
3d
models
and
then
share
them
with
whoever
just
your
an
email
address.
Then
people
and
open
up
their
browser
and
take
a
look
at
it
and
yeah.
C
This
is
just
DC,
resistivity
and
yeah,
so
the
the
objects
that
you
can
display
in
here
are
pretty
much
the
same
as
what
Rouen
just
showed
through
the
OMF
Docs.
So
we
have
like
triangulated
surfaces
lines
basically
and
points,
and
you
can
also
have
also
have
another
example
here
you
we
also
have
volumes,
so
this
could
be
useful
for
for,
like
modeling
results
as
well.
In
addition
to
data.
C
Yeah-
and
so
these
are
just
a
couple,
other
examples
of
projects
that
Paul
did-
these
are
from
his
more
recent
Bangladesh
work
and
yeah
I
again.
These
are
just
like
reasons
to
be
lines
with
wells
and
maps
on
the
top,
so
you
can
kind
of
see
where
they
are
so
yeah
I
don't
know
I
feel
like
this
is
actually
a
pretty
useful
way
to
share
data
and
I
feel
like
it's
relevant
to
some
Pig,
because
this
is
I.
Don't
know
it's
all
3d
data,
yes,.
A
J
C
Yeah
great
so
so
over
here,
these
the
this
is
like
the
elements
panel.
So
elements
is
what
we
actually
call
these
different
objects
so
like
the
typography
is
an
element
and
the
different
ERT
lines
are
each
an
element.
So
these
are
the
actual
data,
and
so
these
have
meshes
they
have.
This
is
this
is
a
typography,
so
it
has
an
image
on
it
and
just
some
data
and
the
different
ERT
sections
have
resistivity
data
on
them
so
yeah
this
is.
C
This
is
actually
the
the
data
that
is
uploaded
to
view
and
then
what
ron
was
saying
was
these
slides?
So
this
kind
of
lets
you
tell
a
story,
you
can
you
can
sort
of
position
this?
However,
you
think
is
interesting
and
move
around
and
then
you
can
add
a
slide.
I
guess
I'm
not
actually
going
to
do
that
on
this
one,
because
this
is
a
public,
a
public
one,
but
I
can
do
it.
I
can
do
it
over
here.
C
H
Not
really
yeah
I,
don't
know
it
didn't,
do
some
sort
of
interactions
with
the
scene
and
like
slice
it
and
measure
yeah.
It's
in
pieces
also
like
when
you
create
these
slides.
You
can
create
a
bit
of
a
conversation
around
them
in
terms
of
like
you
can
feedback
on
there,
which
is
just
a
comment
thread.
C
G
H
At
this
stage
you
know
not
at
all,
it
will
probably
be
bringing
in
some
sort
of
commercial
model,
but
hopefully
we'll
have
something
for
academic
sort
of
incentives
in
the
same
sort
of
way
that
give
hub
husband
your
on.
So
all
of
the
data
that
you
put
in
here
is
defaulted
to
private,
and
only
when
you
share
explicitly
with
someone
does
it
actually
go
public.
G
H
G
C
Yes,
I
can
I
can
I,
don't
know
just
a
little
bit
on
the
permissions
yeah,
so
the
API
is
open
and
you
actually
you
can't
yet
make
one
of
these
actually
public.
You
still
have
to
be
explicitly
share
with
people
you
want
to
share
with
yeah
anyway.
So
there
is
a
Python
client
now
that
you
can
pick
install.
So
this
is
just
pip,
install
lfyou,
API,
client
and
there's
Doc's
online
LF.
You
don't
read
the
docs
at
I/o
and.
C
So
yeah
the
different,
the
different
element
types
we
support.
If
you
can
see,
see
this
so
there's
points
that
lines,
that's
surfaces
and
surfaced,
grids
and
volume
grids,
and
these
all
can
have
data
on
them.
You
scroll
down
here
well,
I
like
on
the
line
set
segments
and
vertices
just
like
Ron
showed
you
can
now
specify
defaults,
which
is
default
visualization
options,
that's
something
you
couldn't
do
from
the
previous
to
know.
Client.
C
C
Plug
in
my
second
screen,
so
I'm
flying
blind
blind
but
yeah
after
you
cook
install
lfyou,
API
client,
you
can
you,
can
Oh
shoosh
I'm
not
paying
attention
from?
Why
is
this
such
a
silly
demo?
Well
so
yeah
you
can
create
a
session
and
you
can
get
an
API
key.
So
everyone
who
has
a
the
first
thing
you
have
to
do
is
sign
up
for
LF,
you
calm,
but
then,
once
you
do
that
everyone
can
generate
an
API
key.
C
Resources,
or
so
we
have
files.
This
is
like
a
rave
and
images
spatial
get
to
you
elements
and
data
and
textures
and
manifests
so
that's
like
when
you're,
when
you're
actually
uploading
everything
together,
you
have
to
put
it
into
something
called
a
view,
and
that's
that's
what
you
can
actually
visualize
and
grant
permissions
to
see
different
stuff
yeah.
So
you,
let's
see.
C
C
So
this
is,
this
is
just
getting
an
old
steno
3d
project,
so
people
probably
already
have
scripts
to
generate
these
a
few
views
to
know
before
and
you
can
just
call
Stan
Oh
Stan
out
of
view,
and
that
gives
you
a
view.
So
this
is
a
this
is
me,
you
think
that
you
can
actually
look
at
in
the
browser
and
within
that
you
have
your
elements
and
so
that
in
this
particular
view,
there's
only
one
wine
set
elements
and
then
from
there
you
can
look
and
see.
There's
a
couple
data
sets
on
it.
C
C
C
C
H
K
That's
that's
great
and
like
I,
never
ended
up
using
steno
3d
at
the
end,
but
I
was
really
thinking
about
that
and
like
now
let
it
come
up
again
in
Z.
Thank
you.
It's
feels
like
it's.
It's
still
speedy,
like
in
another
format
or
like.
What's
your
take
on
that
the
difference
between
it
with
your
ID
to
decommission
city
knows
PD
in
the
future.
Oh
yeah.
H
K
H
B
No
well
thank
your
own
in
Franklin.
This
is
really
it's
super
exciting
to
see
and
I'm
I'm
excited
about
the
OMF
stuff
as
well
to
see
like
what
what
we
can
do
to
interoperate
and
like
have
simple
measures
actually
be
able
to
be
sent
to
any
of
the
visualization
software,
like
that's,
that's
huge
for
everyone.
I
think.
C
And
I
mean
I
think
that,
like
we
will
be
fairly
early
supporters
of
this.
So
even
if
OMF
isn't
exceptionally
meaningful
to
you,
I
think
like
having
the
same
kind
of
meshes
in
discretize
get
into
OMF,
and
that
means
that
they'll
get
into
lfyou
pretty
soon,
which
makes
it
an
even
more
powerful
visualization
thing.
B
Sounds
good
well,
we
can
be
in
touch
on
slack
there's
a
few
pull
requests.
I,
think
people
are
pushing
hard,
and
so
we
can
coordinate
on
those
and
perhaps
bring
water
for
the
next
release
of
simpe
ties.
Doc
Wow
are
in
reasonable
shape
that
pull
request,
but
I.
Think
too
the
comment
we
look
on
the
website.
It
would
be
nice
to
actually
move
the
website.
You
like
something
like
it
have
pages
on,
so
that
more
people
can
contribute
to
that.
B
C
Should
also
mention
from
the
lfyou
client
side
of
things,
I'm
I'm
on
this
MPEG
slack
and
I'm
sort
of
actively
working
on
making
this
client
library
better.
So
if
you
do
want
to
use
it
and
want
hope
or
start
using
it
and
run
into
problems,
so
I'm
just
like,
let
me
know
that's
kind
of
another
selfish
reason
for
sharing.
This
is
like
just
getting
getting
people
using
it.
Improving
things
from
there.