►
From YouTube: SimPEG meeting April 1
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
A
C
Sure
yeah
nothing
too
much
on
development
side,
mostly
some
stuff
internally
I'm
making
an
app.
So
it
makes
it
a
lot
easier
for
everybody
to
run
simpang
conversions,
just
kind
of
like
a
UI
that
talks
to
our
some
of
our
resource,
computers
in
Saskatoon
and
spins
off
inversions.
So
that's
a
ton
of
fun.
C
I'm,
just
using
a
QT
is
pretty
much
for
the
whole
back
and
the
QT
SSL
stuff
and
some
most
of
its
network
tools
and
yeah,
just
a
little
cutie
interface
and
just
sweet
bus
bus.
In
the
background
on
the
resource
manager,
side,
cool
yeah
that
and
yeah
I
finished
off
the
first
round
of
the
AI
stuff.
That
was
pretty
cool.
D
C
Were
just
trying
to
minimize
I
guess
what
what
time
wasted
on
processing
like
it'll,
try
and
process
files
that
are
obviously
just
disconnected
and
just
yeah
nodes
are
malfunctioning.
So
what
the
problem
was
is
just
through
that,
so
that
we
could
just
produce
a
subset
of
the
data
that
is
going
to
produce
good
enough
data
to
process
not
wasting
time
on
all
these
other
kind
of
nodes
and
yeah.
C
Much
I
went
through
a
whole
body
just
like
grabbed
a
whole
bunch
of
time
series
and
I
made
my
own
classification
step,
so
I
went
through
and
just
like,
this
is
or
I
look
at
it
by
the
by
the
full
period
and
yeah
I'll.
Just
like
this
one
ZM
affected
I
took
data
that
didn't
have
multiple
things
wrong
with
it
in
a
sense,
so
it
was
just
like
this
is:
am
coupling
there's
not
gonna,
be
like
for
contact
or
anything
in
there.
I
just
found
data
that
really
highlighted
those
classifications
and
yeah.
F
C
D
C
Know
it's
pretty
much
yeah
just
classifying
it
as
this
is
am
coupling.
This
is
a
coupling
issue.
This
is
contact
issue
and
this
is
to
learn
not
to
lurk.
Sorry
I,
like
transients
yeah,
just
kind
of
classifications
like
that,
and
then
we
have
like
a
sampling
issue
that
sometimes
comes
up
and
I've
taught
it
to
find
that
as
well.
E
F
E
E
Think
that
when
you're,
when
you're,
like
like
applying
your
training,
sets
like
when
you're
applying
your
classifier,
you
don't
really
know
right
like
a
what's
the
beta
right
like
it
could
include
the
whole
bunch
of
them
together.
So
it's
not
like
a
like,
not
as
nicely
separated
as
oh.
It's
like
this
is
yeah
call
playing.
This
is
like
yeah.
C
So
I
defined
the
ones
that
were
mostly
like
safe
enough
to
say
that,
and
so,
like
I
tried
to
minimize
ones
that
had
combinations
like.
Oh,
this
is
a
sample
issue
and,
and
then
I
was
looking
at
some
of
those
stats.
I
haven't
gone
too
deep
into
it,
but
yeah
like
ones
that
have
say
like
a
sampling
nem.
If
you
look
at
the
scores,
you'll
see
that
it
is
registering
in
both
of
them.
It
just
depends
on
how
you
pick
it.
C
F
C
F
E
F
C
Very
end,
yeah
I'm
kind
of
just
doing
it
most
at
the
end
of
the
day.
It's
just
good
and
bad
and
that's
kind
of
how
I
do
it
like
'I'm
coupled.
I
will
include
that
in
the
good
database,
but
it's
just
for
statistics
like
at
the
end
of
it
all
it'll
say
well,
you
know
we
think
this
has
been
not
48
percent
of
this
data
is
got
'I'm
coupling
in
this,
this
one's
bad
or
this
one
had
that
yeah.
C
C
F
F
F
F
E
Nothing
particular
particular
update
synthetic
sympathy.
I
was
helping
like
I
helped
a
little
bit.
Do
you
know
right
like
a
Ryan
Smith
he's
a
professor
at
Missouri
tax,
he's
one
of
the
alumni
in
the
Rose
Marie's
group,
so
he's
trying
to
use
the
cylindrical
code
to
Ghent
like
it
to
simulate
the
karst
example,
so
they
have
like
what
they
call
totem
system
and
then
I.
E
It
so
he
bought
it
and
they
they
are
basically
trying
to
characterize
where
the
cavities
and
the
characters
of
the
cavity
so
he's
what
he
wants
to
use
this
impact
as
simulator
and
do
a
machine
learning
about
what
it's
like.
So
you
can
generate
a
whole
bunch
of
simulated
data
where
you
have
like
different
size
of
the
cavity
depth,
different
resistivity
and
you.
E
You
learn
that
and
potentially
estimated
what
are
the
depth
of
the
cavity
like
resistivity
of
the
cavity
or
something
like
that
so
yeah
that
seems
fun,
interesting,
applicate,
I
think
people
are
moving
on
that
direction
more
and
more
so
use
like
so
they
recognize.
Oh,
the
field.
Data
is
not
enough,
so
we're
gonna
use
this
in
the
weighted
data
and
then
train
it
and
see
if
you
can
get
the
right
answer.
So
that's
something
interesting
and
the
other
thing
that
I
was
collaborating
with
the
topographic
effects
in
in
the
open,
VM
data.
E
So
you,
yes,
yes,
sky,
I,
think
it's
Paul,
Paul
Bedrosian
and
one
of
his
colleague
named
Katrina,
so
they
got
the
airborne
data
from
San
Andreas
Fault.
So
this
is
an
interesting
location
because
is
quite
conductive.
So
even
if
you
got
like
a
big
topography,
if
you
are
in
a
relatively
resistive
area,
so
it's
not
that
much,
but
if
you
are
actually
in
their
in
their
conductive
area,
so
that
topography
effect
could
kill,
kill
your
1d
inversion.
So
this
is
an
interesting
example
that
they
kind
of
faced.
E
E
A
C
E
A
Dates
or
anything
yeah.
B
Excited
about
this,
it's
nice
to
be
able
to
invert
a
tree
on
rotated
meshes,
and
you
know
both
grab
and
mag,
quit
pretty
quickly
nice
and
I'm
working
on
a
directive,
because
right
now,
like
the
NDI
code,
that
I
was
running,
you
know
Cartesian
first,
the
linear
part,
and
then
it
would
be.
You
know
a
whole
bunch
of
scripts
which,
to
a
spherical
so
and
I'm,
trying
to
write
the
directive
that
will
do
the
vector
inversion.
All
in
one
go
and
just
do
this
switch
internally.
So
we
don't.
B
D
That
yeah,
it
started
to
get
to
write
the
extended
abstract
or
the
SVG,
so
gwb
will
have
a
workshop
there
and
we
have
an
opportunity
to
present
at
least
a
couple
and
I'm
going
to
talk
to
the
new
gal.
She
it's
a
levy,
I
think
it's
her
name
anyway.
She's
taken
over
from
Linda
for
I've
got
a
schedule
phone
call
with
her
tomorrow.
I
just
want
to
find
out
what
this
workshop
has
or
that
presentation
session
and
see
what
kind
of
flexibility
there
is.
D
But
at
this
moment
certainly,
we
can
present
one
paper
that
just
talks
about
the
GWB
project
as
a
whole
paper
that
I'm
trying
to
put
raph
together
on
actually
has
you
know
the
role
of
open
source
in
capacity-building.
So
we've
got
actually
a
really
really
nice
story
to
present.
From
from
that
perspective,
I'm
going
to
talk
to
this
gala
tomorrow,
but
I
think,
there's
perhaps
and
another
possibility
to
actually
talk
about.
D
You
know
the
developments
of
codes
and
simpe
algorithms
for
doing
simulation
and
conversion
and
really
kind
of
taking
this
down
to
another
level
so
effectively
there.
There
might
be
like
three
levels:
what
like
ones
just
at
a
at
a
high
level
for
what
the
project
was
and
then
the
second
one
really
focusing
on
your
we're.
Open
source
plays
a
role,
and
it's
everywhere
from
you
know
that
you
know
the
slides
and
the
course
website
and
capturing.
D
You
know
the
videos,
as
well
as
all
of
the
the
jupiter
notebooks
and
case
histories
and
how
that
really
is
kind
of
the
glue
that
allows
people
to
communicate,
and
you
know,
reproducible
results
and
all
this
stuff.
So
that's
sort
of
that
immediate,
but
there
is
potentially
I
think
another
level.
That's
farther
down.
That
really
is
talking
about
build
the
codes
that
are
developed
and
simpe
and
distribution
of
them.
A
Yeah,
the
couple
updates
that
I
have
got
an
email
from
Xander
who's,
that
Aurora
geosciences
they're
wanting
support
for
getting
e.l.f
inversions
in
sim
thing.
So
again,
somebody
else
to
once
you
know
natural
fields,
natural
fields,
see
em
up
and
running,
and
you
know
few
more
examples
and
things
like
that:
they're
willing
to
actually
pony
up
some
money
for
that.
So
I,
don't
know.
A
If
there's
anybody
who
sort
of
actually
in
a
position
where
you'd
like
to
sort
of
take
on
a
bit
of
contract
style
work
like
we
could
probably
scope
that,
so
that
it's
something
where
you
are
just
responsible
for
generating
a
few
examples,
because
I
mean
there's
enough
of
us
who
want
to
do
a
push
on
on
empty
and
natural
fields
like
you,
don't
have
to
be
rewriting
everything,
but
if
somebody
sort
of
once
yeah
a
contract
style
position,
that's
something
we
can
absolutely
do
I
think
and
if
not
I,
think
there's
still
like
opportunity
to
connect
with
them
and
loop
them
in
and
actually
try
and
sort
of
get
them
engaged
and
actually
creating
examples,
which
would
also
be
a
good
thing
and
I.
A
They
so
they
didn't
actually
provide
a
dollar
value,
but
they've
asked
they'd
like
to
know
if
they
could
pay
somebody
to
get
PLF
data
inverting
insane
things
so
I.
Imagine
like
a
contract
style
position
where
you
could
build
by
the
hour
kind
of
thing,
and
we
can
absolutely
like
schedule
a
call
to
get
a
sense
of
that.
A
If
that
is
of
interest,
I
think
we
should
schedule
a
call
with
with
them
probably
either
way,
but
it
would
be
good
if,
if
we
can
sort
of
figure
out,
is
this
something
that
somebody
wants
to
sort
of
put
on
their
plate
a
bit
more
formally?
Or
is
this
something
that
we
just
kind
of
work
with
them
informally,
because.
D
It
at
some
point
we
might
want
to
scope
out
beforehand.
You
know
what
the
functionality
and
the
capabilities
of
our
empty
code
are
going
to
be
because
you
know
for
the
e.l.f
data,
it's
just
it's.
Basically
tipper
data
like
sandwich
over
my
going
to
field
at
the
particular
location,
and
certainly
all
thoughts
about
what
our
empty
code
would
be,
would
definitely
have
that,
as
well
as
whatever
combinations
of
transfer
functions
you
might
want
to
think
about,
as
well
as
they're,
not
measuring
electric
field.
D
So
it's
basically
just
is
that
temp
type
of
stuff,
but
depending
on
where
you
put
your.
D
But
I
don't
know
it
might
just
be
worthwhile
for
us
as
a
group
just
to
kind
of
scope
out.
You
know
what
we'd
like
to
see
in
that
empty
world
and
as
I
said
last
week,
dandy
Quan
was
also
kind
of
interested
and
yeah
just
sort
of
figure
out
where
we
might
want
to
go
and
who's
interested
and
then
see
how
you
know
Aurora,
but
certainly
they
would
be
probably
very
good
from
the
point
of
view
of
providing
datasets
and
they
might
also
be
a
really
good
group
to
test
things.
D
B
C
I'm,
sorry
they
reached
out
on
discourse
there
and
they
posted
a
line
of
data.
I
was
gonna,
get
to
it.
I
did
and
I
reply
back
saying
that
like
well
yeah
this.
This
will
be
easy
because
I've
already
been
doing
some
tipper
fort
modeling.
It
would
just
need
to
build
a
loader
for
their
data,
and
that
was
where
it
just
kind
of
fell
off
so
I'm.
Judging
from
this
I
I
feel
that
there's
nobody
there
that
has
Python
experience
is
what's
happening.
Okay,
they
need
yeah.
C
They
just
need
someone
to
really
build
a
loader
and
what's
interesting.
There
is
I'm
kind
of
in
the
process
of
doing
that
too,
with
our
data
that
we
just
had
got
back
from
South
Africa,
so
I'll
be
building
pretty
much
nioh
in
a
sense
that
is
similar
because
they're
not
using
EDI
files,
either
they're
just
using
a
text,
so
I
guess,
I,
yeah
I
would
be
happy
to
rehash
my
code
if
they
really
needed
it
kind
of
thing
is.
C
It's
kind
of
weird
but
I
think
at
this
point
like
it
that
wouldn't
matter
it's
gonna,
be
like
it's
open
source.
So
it's
not
like.
We
were
trying
to
hide
it.
Anyways
and
yeah
code
is
the
code
and
it'll
run
run
the
same
way
and
I'm
loading
the
file.
It's
pretty
much
similar,
how
they'll
be
doing
it,
so
it's
just
yeah
getting
their
format
and
putting
it
in
yeah.
F
F
C
A
Can
think
through
that
John
to
is
like
potentially
even
laying
out?
Basically,
what
are
the
steps
that
they
would
need
to
take
I
mean
if
there's
already
sort
of
a
template
that
they
could
build
from
and
then,
if
none
of
us
have
the
bandwidth
to
really
do
that.
You
know
a
couple
of
us
basically
helping
like
Shepherd
them
figuring
out
how
to
just
write
a
simple
Python,
loader
file,
yeah.
B
A
Might
not
be
a
bad
idea
is,
like
you
know,
to
get
some
investment
from
from
them
on
the
table.
If
it's
not
cash,
it's
time,
sorties
time,
yeah
yeah.
Would
it
be
worse
and
I
think
this
is
sort
of
on
the
radar
but
having
a
just
dedicated
empty
natural
fields,
meeting
to
started
just
start
planning
some
stuff
like
either
related
later
this
week
or
early
next
and
sort
of
cash
out
a
bit
of
a
plan
because
it
seems
like
this
is
something
that
is
kind
of
burning
for
a
number
of
us
and
so
yeah.
A
D
A
We
started
that
Google
Doc
that
I
sent
around
for
the
Jupiter
planning,
but
it
might
be
worth
actually
turning
that
into
like
a
bit
of
a
discourse
post.
Would
that
make
sense?
Do
you
guys
think,
and
we
can
iterate
on
that
a
bit
just
so
it's
a
bit
more
in
public
and
we
can
even
point
some
folks,
like
Aurora
tickets,.
A
I
can
start
something
like
that
and
I
think
you
can
do
it
like
a
wiki,
so
anybody
can
sort
of
edit
like
the
main
text
so
that
we
don't
just
end
up
with,
like
a
shining
thread,
I'll
try
that
if
it
becomes
too
clunky,
we
can
just
move
back
to
a
Google
Doc
and
just
like
leave
a
Google
Doc
link
and
the
discourse
post.
I
think
that's
also
totally
fine.
What.
B
A
So
with
the
Jupiter
research
project,
we
need
to
deliver
well,
we
need
to
deliver
basically
scientific
use
cases
and
so
I
pitched
the
MT
I'm
gonna
call
Bedrosian
actually
and
get
some
data
from
him
to
play
around
with,
but
basically
what
I
need
to
get
out
of
this
for
the
Jupiter
means
the
earth
project
is
some
like
a
concrete
sort
of
development
tasks
and
or
things
that
are
somewhat
unique
to
the
Jupiter
ecosystem.
A
So,
for
example,
we
we
now
have
a
Jupiter
hub
developer
sort
of
he's.
Well,
he
he's
on
two
days
a
week
at
this
point
on
the
Jupiter
Maisie
earth
project,
and
so
one
of
the
things
that
we're
looking
at
trying
to
get
an
understanding
of
is
a
problem
like
MT.
Does
it
make
sense
to
be
running
it
on
the
cloud,
and
so
we
have
access
to
cloud
resources
and
so
playing
around
then
with
simpang
on
the
cloud
and
sort
of
comparing
that
then
also
to.
E
There's
a
Sims
of
two
levels
right
like
they're,
about
the
higher
level
to
handle
the
hog
hunter
and
the
crop
Claude
resources,
but
you
also
have
a
lower
level
problem
in
the
code,
how
that
could
be
like
paralyzed
or
structured
to
be
run
and
in
that
system?
I
guess,
is
that,
like
you
like?
Are
you
trying
to
handle
both
of
them
in
Z
or
like
you're,
more
on
to
the
upper
level
structure?
E
A
I
guess
my
interests
are
more
on
the
like.
Well,
there's
a
bit
of
both,
but
on
the
simple
side
of
things,
that's
where
I
plan
to
be
investing
effort,
but
then
in
in
conjunction
with
working
with
folks,
who
are
on
the
Jupiter
club
side
of
things
and
can
help
me,
like
anything,
elbows
figure
out.
What's
the
best
way
to
paralyze
and
all
of
that
sort
of
stuff.
F
A
With
tasks,
we
should
be
portable
to
both
locations
pretty
easily,
but
you
know
that's
something
that
we
say
verbally,
but
you
never.
You
never
really
know
until
you
actually
start
hitting
on
these
things.
So
yeah
does
that
answer
your
question.
B
B
E
Just
like
yours,
who's
really
championing
this
empty
code.
I
got
because
there
are
two
levels
right
like
there
could
be
a
level
of
like
computation
like
Oh
Howard
parallelizing
this
and
make
you
really
efficient.
That's
one
side
but,
on
the
other
hand,
side
like
like
in
community
to
make
sure
the
physics
works,
and
why
that
the
boundary
conditions
right.
There
are
lots
of
details
about
like
setting
things
really
right
and
then
test
like
with
like
more
realistic
data
such
that
IO,
that
that
could
really
works.
Do
we
have
like
a
I?
E
D
D
Oh,
it's
coded
how
to
work
with
the
boundary
conditions
and
yeah
get
into
the
physics,
so
I
think
from
the
point
of
view
of
working
and
developing
a
code
so
that
we
can
change
the
physics
and
you
know
the
boundary
conditions
or
things
like
this
I
think
Joe
has
the
potential
for
playing
a
real
central
role,
and
since
he
at
this
point
is
you
know
really
the
only
person
who
has
got
a
mandate
to
kind
of
work
within
that
same
peg,
environment
and
get
pay
for
it.
You
know
it's
kind
of
the
kind
of
words.
D
So
I
think
that
if
the
rest
of
the
group
can
we
can
kind
of
get
together
and
really
decide,
you
know
what
we
what
we
want
to
have
as
a
final
product
and
try
to
see
what
bits
and
pieces
are
around,
because
everybody
perhaps
knows
something
that
is
going
to
be
abused
but
actually
to
have
Joe.
Being
you
know
a
real
key
point
in
this
at
this
point
from
the
people
that
we
have
here,
I
think
that's
sort
of
the
logical
way
before
that.
B
Joe
nice
plan
right
now
is
to
is
to
have
our
respective
companies
pony
up
men
our
of
our
hours,
to
do
this,
but
more
on
that.
You
know
on
a
usable
usable
side,
so
Joe
can
handle
like
being
the
testing
and
all
that
stuff.
That's
great
and
then
John
I
can
look
at
more
of
the
the
usability
of
a
paralyzation.
What
would
Lindsey
and
what
not
yeah
yeah.
E
D
You
know
it's
on
this
screen
at
this
point.
You
know
a
primary
mandate
and
then
trying
to
fit
in
you
know
simpe
development
or
empty,
or
something
like
this
as
it
coincides
with
what
their
projects
are,
so
that
you
can
kind
of
kill
two
birds
with
one
stone,
but
nobody
that's
got
really
time
to
say.
Okay
here
is
my
main
project,
so
I
think
that's
where
you
know
the
benefit
of
having
Joe
is.
D
A
In
the
meantime,
I
can
I'll
try
and
basically
put
what
notes
I've
taken
into
a
discourse
post
and
we
can
sort
of
try
and
work
asynchronously
so
that
we've
got
a
bit
of
a
document
to
start
from
that
week.
Does
anybody
have
a
preference
on
day
we
put
the
13th
I.