►
From YouTube: SimPEG Meeting - Oct 31th ~~~ boohooo!
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).
B
C
E
A
D
B
G
D
B
D
F
D
Thing
yeah,
so
if
you
look
myself
above
and
then
I
merely
to
chat
with
Glenn
who's
Diaz-
and
this
is
there
were
the
interested
in
potentially
you
can
submit
for
every
similarity
see
data
as
well.
So
I
quote
that
quality
yesterday
for
just
sewer
to
see
examples
just
to
show
that
like
hey,
it's,
it's
not
marinara,
we
can
actually
work
previously
and
one
of
the
things
that
I've
hunted.
B
B
E
B
D
D
I
D
Yes,
there's
nothing
to
you
evolve
with
even
apply
the
Oslo
Cornell
next
like
a
bit,
but
if
we
kind
of
compute
a
few
more
examples,
especially
like
3d,
BCC
I,
think
that's
something
we
should
try
for
eventually
for
all
of
the
methods
is
takaka
like
here's
a
practical
example
right
and
with
that
said,
the
computations
that
we
can,
we
should
have
got
up
to
gsi,
yeah
and
so
even
was
her
names
in
there.
It
now
I
think
if
it's
visible,
we'll
get
on
anymore.
D
B
D
J
J
J
B
J
J
K
B
J
G
J
J
Yeah
plus,
like
best
and
I,
think
what
I
like
there's
gonna,
be
missing.
This
piece
was
slave
and
soft,
her
side
and
I
don't
know
why
people
say
I'm,
not
sure,
but
they
were
saying
I
hope
now
we're
at
the
stage.
We
can
form
all
I.
Remember
the
most
two
types
of
data,
there's
probably
true,
but
the
missing
piece
is
communication
between
like
different
software's
and
how
to
like.
How
are
you
really
integrated
but
I,
really
enough
communicate
different
physics
and
also
the
canonical
versions
of
your
school
knowledge?
J
J
D
D
D
B
J
J
J
D
It's
like
really
I,
which
is
20
awesome,
like
it's
good,
to
see
that
things
are
under
there,
but
we
also
need
a
way
to
say
like
hey.
This
is
in
progress.
Looking
at
it,
Chris
is
like
this
I'm,
not
gonna,
jump
in
and
start
helping
review
sort
of
solo
yeah
or,
if
there's
a
specific
questions
and
things
like
that,
they
have
yeah.
So
things
like
there's
the
the
MIP.
D
I
B
I
D
I
D
B
Yeah,
so
we
could
set
it
up
if
you
guys
want
basically
creating
well.
Let's
look
at
not
looks
for
origin
tools,
but
the
neat
part
is
that
you
can
start
like.
We
could
basically
create
a
project
for
physics
or
something
like
that
and
then,
after
that,
what
I
own
is
pretty
pretty
light,
but
you
can
have
in
progress
and
you
can
move
these
guys
and
you
review
and
want
review
on
that
right.
J
B
D
D
B
D
D
D
D
D
D
C
D
C
K
D
B
J
A
J
B
Yeah,
it's
pretty
big,
but
I
can
I
can
transfer
most
of
it.
How
is
your
drug
box
in
your?
Can
you
hold
yeah
I
finally
bought
the
terabyte
goes
like
I'm,
tired
of.
A
B
On
if
this
makes
no
sense
of
it,
it's
only
sitting
locally,
but
it
actually
that's
the
thing.
We
need
to
figure
out
what
that
is.
We
have
a
lot
of
case
studies
and
you
want
to
start
bringing
them
in
you
know
so
Sarah
can
do
the
cleanup
and
Chris
had
some
on
this
computer
that
it's
got
scattered.
We
need
to
bring
them
all
into
like
a
UVC
umbrella
or
whatever
happens
to
UVC
later
they
have
access
to
what
me
what
we
have.
H
D
Itself
before
like,
if
you
just
want
to
lick
the
wave
reason
right
now,
we
just
have
stuff
up
there
that
you
want
to
hit
and
download.
That's
really
you
saying
you
didn't,
and
so
yeah
I
would
think
like
a
first
pass
thing
would
just
be
like
get
a
job.
F
Before
anybody,
the
details
of
a
set
of
the
code,
I
kind
of
only
give
you
guys
five
minute
reminder
of
this
code.
Actually,
does
you
guys
understand
why
so,
let's
say
you
thought
sort
of
true
erm
response
on
sort
of
a
log-log
plot
and
would
do
something
like
this,
but
I
found
that
over
the
range
of
time
channels,
which
we
measure
time
convenient
data,
we
have
an
approximation
to
that
decay,
and
this
lets
us
predict
the
fields
according
to
a
linear
equation.
So
where
n
is
the
model
parameters,
maybe
I
don't
know?
F
F
Whether
I
want
I
want
to
measure
like
H
or
db/dt,
it's
going
to
populate
a
very
sparse
matrix
T.
So
when
you
look
into
this,
you
realize
it's
a
linear
problem
and
we
have
a
framework
for
this.
However,
the
more
time
channels
you
use
the
more
difficult
it
is
to
store
it's
noisy.
So
with
the
current
somatic
framework.
I
can't
distort
the
linear
mockery.
F
F
It
was
really
important
for
me
and
probably
for
users
getting
used
to
simpang
that
if
I
made
some
kind
of
class
and
I
built
it
off
of
some
super
class,
all
of
the
functionality
that
I
want
for
this
needs
to
be
in
this,
but
at
the
same
time
I
don't
want
redundant
functionality
or
redundant
operations
that
I'm
allowed
to
call
at
Dover
go
to
this
class.
So
I
put
the
DRM
in
a
different,
a
different
folder,
so
that
was
really
importantly.
One
example
is
the
right
hand,
side
I,
have
some
magic
act
as
a
problem.
F
I'll
have
some
transmitter,
but
there's
no
need
to
compute
right
hand
side.
So
it's
what
we
went
through
with
that.
The
other
thing
which
maybe
is
a
third
point,
is
if
we
have
some
kind
of
time
to
medium
problem.
We
typically
have
sort
of
en
response
that
the
vrm
does
this.
The
whole
point
of
developing
all
of
this
was
that
we
wanted
to
invert.
Just
these
tiny
channels
predict
the
data
everywhere
and
then
subtract
X.
F
This
will
basically
it's
the
it
sort
of
just
using
a
projections
that
you're
only
using
the
cells
that
have
defined
as
if
at
the
are
a
brief
response
to
the
narrator.
It's
only
for
that
layer
yeah,
you
can
specify
it
and
I
know
that
we
have
an
active
cell
projection.
I'm
I
have,
for
various
reasons,
I'm
using
both
and
a
like
a
topography
projection.
It's
just
a
way
of
reducing
the
size
of
a
is
in
certain
situations.
The
way
I
was
defined
things
in
the
code.
F
F
F
D
Yeah,
it's
it's
just
a
question
of
how
you
interact
and
I,
don't
even
know
on
your
floor.
Questions
I!
Think
that
this
clarifies
position
a
quick
look
at
Connie,
Lee
instantiate,
your
your
problem
class,
so
I
think
the
way
that
the
math
Nathan
Scott's
can
just
learn
all
of
the
very
standard
way
that
we
always
prĂłxima
into
a
problem.
So
it
doesn't
have
to
be
treated
separately
and
everything
like
that.
So,
even
though
it's
it's
own
distinct
style,
it
can
still
be
treated
in
the
various
third
place.
F
B
E
N
K
J
J
F
F
B
J
B
F
B
A
B
F
F
F
You
would
need
to
do
something
about
it,
so
that,
if
you
had
some
kind
of
transmitter
receiver
position
here,
you
can't
be
for
modeling
cells
that
are
there
for
multi
locations
that
are
inside
cells
right
potentially.
So
we
need
something
for
that:
let's
sort
of
power,
but
I've
defined
a
physical
property
that
I'm
calling
you
see
and
I
have
like
CC
map
and
see
derivative,
so
I'm
like
I'm,
using
the
same
the
same
as
much
of
this
framework
as
I
can
from
the
base
problem.
F
F
Over
long
and
tau,
1,
tau,
2
and
some
other
stuff
and
I
was
able
to
make
this
amalgamated
like
viscous
property,
it's
Delta,
Chi,
so
sort
of
the
amplitude
of
the
discuss
viscous
effect
over
all
frequencies,
/
well
I
mean
County
over
tau
1.
It's
just
sort
of
this
amalgamation
of
a
few
parameters
describing
magnetic
Wisconsin
did.
F
G
F
H
F
D
D
F
Great
yeah,
but
for
the
linear
vrm
collapse,
it's
one
of
the
first
things.
That's
defined
it's!
It's
yes,
define
as
a
property,
an
invertible
property.
Yes,
the
physical
property
that
is
associated
with
that
feeling
it
with
some
commenting,
and
if
you
understood
how
to
work
with
the
sim
petticoat,
it
should
be
pretty
obvious
that
that's
how
I'm
dividing
my
physical
property,
like.
F
D
Yeah
and
I
think
one
of
the
things
that
I
certainly
changed
when
the
world,
but
when
exploring
an
API
I,
don't
read,
I,
never
create
an
object
and
I've
got
to
have
no
yo
menu.
If
I
see
see
and
like
okay,
what
I
see
so
that,
like
just
sort
of
helping
facilitate
that
sort
of
interaction,
I
think,
is
actually
important
in
terms
of
making
sure
that
very
attractive.
F
B
F
F
Okay,
so
that's
pretty
much
all
I
can
say
about
the
background
physics,
but
yeah,
that's
sort
of
yeah.
These
sort
of
details
that
are
pretty
important
is
the
linear
problem,
but
we
can't
just
use
preexisting
that
your
framework,
because
we
want
to
store
TNA
separately,
it
supply
the
main
problem.
You
don't
want
to
use
a
bunch
of
redundant
functionality
and
some
of
the
stuff
with
the
receivers
is
different
and
then
you're
gonna
want
to
invert
a
subset
of
the
time
channels.
F
F
This
is
all
of
the
the
geometric
relationships
between
the
transmitter
in
the
cells
and
the
receivers
in
the
cells
and
T
I'm,
calling
it
the
characteristic
decay
majors,
where,
if
the
magnetization
is
just
on
purely
time,
dependent
function
times
a
small
parameter,
then
you
have
this
sparse
sort
of
block
diagonal
matrix,
that's
populated
with
this
root
decay.
Behavior,
like
the
time
dependent
function
so
things
we
use
for
all
yeah
same.
C
F
F
F
So
there
is
a
sort
of
refinement
just
sensitivities
for
that
cell.
It
breaks
the
salary
to
a
bunch
of
small
volumes
for
recompute
the
sensitivity
and
then
yeah.
So
there
is,
you,
don't
have
to
put
them
in
there
KW
arguments.
This
looks
sort
of
a
standard
that
gives
you
a
fairly
good
hair,
the
accuracy
we
can
change
it.
You
can
also
put
it
in
this
Apocrypha
mapping
or
you
can
not
put
it
it'll
just
be
detonated
in
natures,
and
then
we
have
our
seed
mapping
that
either
starts
out
as
not.
F
F
Right
so
the
RM
is
a
source.
It's
a
line,
current
source
we
put
in
the
location,
the
notes,
the
current
and
then
waveform,
so
pretty
similar
flavor
till
you
have
more
time
to
medium
problem
and
then
you,
you
top
of
a
survey
with
the
transmitters
and
then
you
you
can
either
choose
to
or
not
choose
to
set
the
range
of
time
channels
that
you
want
to
invert.
So
if
I
was
to
do
that
later,
maybe
I
only
want
to
invert
everything.
That's
after
one
milliseconds
cuz,
that's
where
CPR
and
response
dominated
you
can
set
back.
F
You
in
Japan,
through
broader
than
interval,
is
that
I
set
it
up
to
be
versatile
in
a
sense
that
what
if
somebody
was
a
crazy
person
and
decided
that
they
wanted
a
whole
bunch
of
different
transmitters
with
different
waveforms
and
different
receivers,
and
they
wanted
to
measure
different
rate
different
time
channels.
It
has
to
be
able
to
expect
that
scenario,
so
this
was
I
mean
if
I
put
in
the
individual
time
channels,
then,
like
me,
I,
don't
know
I'd
be
looking
for
for
matches
times.
The
interval
make
more
sense.
F
You
just
put
in
two
numbers,
as
element
I'm
gonna,
to
find
all
of
the
places
where
I'm
getting
responses
between
this
place.
Look
like
that!
What's
your
dad's
at
two
points
or
you
will,
he
will
give
the
difference
between
fields
and
he
predicted
there's
that
piece
of
deep.
There
would
be
some
kind
of
projection
matrix
that
we
tell.
F
The
thesis
that
whole
collapse,
anything
in
this
framework
I
can't
lie
so
I
only
want
to
be
doing
whatever
you
interacted
and
putting
it
in
Savanna
conversion
framework,
and
my
solution
to
that
is
that
I
take
the
radiator
five
channels,
I
want
to
invert,
and
so
D
predicted
and
D
observed,
and
all
that
stuff
will
basically
map
fields
to
data
using
the
projection
matrix.
It's
exactly
how
all
the
other
goes
at
work.
This.
F
J
F
A
song
part
yeah,
so
if
I
check
fields
so
in
this
case,
I
got
I,
have
three
transmitters
each
with
one
receiver
and
each
of
the
measuring
at
three-card
channels,
right,
negative,
4
to
negative
2
so
about
990
total
data,
but
then
I
defined
that
I
wanted.
Also,
these
active
is
active
time
interval
here.
Is
it's
only
going
to
take
the
middle
of
this?
So
if
I
know
do
you
predict?
Did
it's
only
this?
H
F
This
is
just
a
method
that
was
defined
in
the
base
or
a
class
I
didn't
take
this
so
I'm
trying
to
use
any
any
any
superclass
and
I'm
using
that
I'm
using
to
face
the
vrn
classes.
I
want
to
be
able
to
use
all
that
much.
You
know
how
to
have
no
function,
but
you
can
do
the
same
thing.
Rice,
you
can
pair
to
survey
with
the
problem.
Predict
fields,
unpaired
all
that
all
that
stuff
put
it,
and
so
then
the
last
thing
that
they
didn't
want
to
go
over
in
listening
has
any
questions.
F
F
F
F
Inner
vertical
properties
for
the
problem,
so
I
made
a
sort
of
a
secret
after
peak
underscore
a
and
I
call
the
property
egg.
So
I
was
like
problems
on
a
if
I
set
it
Eric.
Sorry,
it's
not
set
yet
it'll
check
if
I
pair
the
problem
with
a
survey
and
it
ex
corde
something
we
don't
delete
it
unlocks.
Where
do
you
think
that.
A
B
F
B
Yeah
and
then
and
then
509
507,
you're,
alway
always
doing
a
projection,
you're,
always
driving
in
your
full
space
and
your
project,
and
you
think
II
could
only
do
that
at
the
end
and
end
up
everything
like
you
don't
need
to
do
that
at
that
operation
here,
every
time
you're
doing
it
here,
you're
doing
it
at
205,
489
yeah,
you
don't
need
it
there,
because
you're
ready,
reducing
your
space
or
work
in
the
reduced
space.
All
the
way
inverting
reduce
finish
and
then
the
end.
F
B
Q
that
M
doesn't
have
to
be
full
space
and
could
be
only
the
size
of
your
a
well.
That's
what
I
mean
so
you'll
skip
yourself,
always
skull
this
guy.
This
mapping
do
the
transpose
every
time
you
do
them
a
to
respect
their
brother
because
you
don't
need
to
do
it
I
think
what
combined
with
heat?
If
you
a
single,
no,
you
don't
need
to
project,
you
don't
need
to
reproject
every
time
you
only
work
in
the
reduced
space.
You
do
everything
your
inversion,
your
for
and
everything
in
the
reduced
space
and.
B
B
You're
doing
more
operations,
and
you
need
to
doesn't
need
to
be
your
space-
is
only
active
on
your
active
cells.
N
B
But
that's
what
I'm
saying
yeah,
no
the
flooding
at
the
flooding
action!
You
only
do
it
at
the
end
when
you're
done
right,
so
you
you,
don't
you
don't
need
to
do
it
inside
here?
This
can
be
much
more
efficient.
If
you
don't
need
to
always
call
the
transpose
and
bring
it
back
and
bring
it
back
every
time,
yeah.
B
J
F
B
M
B
M
M
B
B
All
night
it's
on
my
wrench
for
the
for
the
grout.
It
will
be
Gobi
brand
rather
than
once
I
finish
the
test.
Yes,
it's
the
whole
thing
that
it's
there's
a
difference
between
Michal
sell
and
when
you
call
I
grew
at
CC,
something
like
that.
So
style
is
always
not
if
you
know,
there's
no
origin
honor,
but
grit.
Cc
is
origin
to
me.
So
it's
just
there
yeah.
F
Yeah
the
bottom
line
when
I
wrote
this
we're
gonna
store
a
and
we're
gonna
store
to
you,
and
if
you
were
to
change
something
like
the
a
few
cells,
I'll
change,
the
pendency
of
a
for
any
kind
of
reason-
and
all
you
would
be
all
you
have
to
do-
is
no
problem,
and
it
would
recognize
that
you
came
something
and
it
would
reform
that
a
20-story
thanks.
So
maybe
this
is
a
good
segue
into
the
properties
thing
that
I.
F
It
could
be
something
that
has
sort
of
you
know
you
don't
bother
for
plenty,
what
they
it's
a
default
value,
but
if
you
wanted
to
even
Senate
when
you're,
initially
you're
giving
an
instance
this
thing
it
won't
call
any
of
the
stuff
below
sort
of
independent
of
that
it
will
the
way
that
I've
written
this
little
script
it
will
yeah,
won't
take
any
two
properties
into
account.
Then
we
need
to
find
a
property.
F
Basically,
I
want
to
make
it
so
that
if
I
I
want
to
call
this
variable,
I
want
to
initially
maybe
some
conditions.
Maybe
I
want
to
hurt
that
I'm.
Getting
this,
this
property
or
maybe
I,
need
to
compute
it
or
what
I
did
before
is
recognize
that
if
I
change,
this
topography
object
that
need
to
recompute
the
eight
liters
before
I
give
it
to
you.
F
F
Some
attribute
we
even
like
the
option
of
eating
aesthetic
or
give
it
a
default
value,
but
we
initialize
the
class
like
create
an
instance
of
the
class,
and
then
we
want
to
basically
be
able
to
control
the
getter.
So
when
I
go
instance
dot
that
property,
it
will
call
everything
that
I
put
under
here.
It'll
it'll
carry
that
out,
put
a
limitations
on
what
I
want
to
set
it.
So
numpy
or
anything
I,
don't
want
to
be
able
to
set
it
as
a
float
or
a
string.
H
F
F
F
F
F
O
F
I
D
Doing
well
it
sort
of
augment
submitted
the
stutters
and
also
it
makes
it
a
little
more
clear
wary
for
sometimes
are
so
setting
up
email
value,
so
saintly
conductivity
should
not
people.
It's
it's
a
little
more
transparent.
We
do
that
through
that
a
property
library.
It's
because
of
the
way
it's
documented,
unlike
its
input
as
a
few
words.
Clearly
its
input,
that's
an
argument
when
you
can
finally
variable
under
the
hood.
It
does
all
of
this
yeah.
That's
what
I
figured,
but
in
a
little
bit
warm
a
transparent
wait.
D
M
D
B
F
And
well
you're
defining
the
property
but
didn't
through
this.
Are
you
also
defined
anything
getters?
There's
the
setters
and
Latinas
yeah?
So
you
because
there's
I
thought
sort
of
here
so
there's
like
a
is
set
and
when
you
change
the
things
they
don't
mind
after
false
flag,
there's
not
sure
how
it's
going
to
line
up
with
the
properties.
I!
F
F
D
D
Okay,
but
yeah
exactly
so
what's
helpful
here
is
that
like
I've
made
it
inline
comment
yep,
and
you
said
you
can
do
this,
nothing
see
we
think
and
so
like.
If
you
have
comments
or
questions
of
without
giving
it
in
place
and
then
what's
nice
is,
if
you
actually
do
going
to
change,
and
this
comment
posts
or
this
your.
F
H
D
F
F
F
D
D
So,
like
I
think
in
order
to
bring
in
my
new
contribution
it's
going
to
take
a
few
passes
and
I'm
sort
of
wondering
how
we
want
to
structure
that
Sidney,
like
especially
if
we
need
contributors.
It's
really
important
that
it's
clear
that
what
we're
trying
to
achieve
in
each
stage
and
if
you
get
a
four
different
reviews
and
each
one
sex
music
stuff.
When
you
go
on
the
line.
That
would
be
really
frustrating.
D
D
D
D
D
B
F
D
N
D
Yeah,
and
so
that's
weird,
like
a
review,
I
think
it's
for
the
hopefuls,
because
we
will
get
a
few
perspectives
on
it
and
try
and
decide
if
it
is
a
community
where
we're
trying
to
go
so
it's
not
necessarily
that
they
call
them.
It
focused
currently
means
wrong.
It's
not,
but
as
much
as
we
can
trying
to
like
sort
of
point
new
contributions
in
the
right
direction,
that's
really
help,
and
so
that's
really.
A
huge
part
of
the
purpose
of
the
review
is
to
really
help
get
some
feedback.
D
D
B
H
B
That
J
Becky
do
they
go.
That's
the
properties,
your
eye,
I
guess
properties
is
like
throughout
right.
We
need
kind
of
a
need
to
review
my
codes
and
everything
like
yeah
and
then
we
can
go
optimize
the
last
pass.
Then
you
could
do
an
optimization,
like
I
figure
out
where
we
can
flee
yeah
organization.
B
K
B
K
B
B
D
F
Get
it
the
ultimate
goal
that
I
have
for
agony
right
stuff
is
I
want
to
predict.
The
rme
predict
the
time
Canadian
code,
add
the
time-domain
p.m.
data
to
the
BRM
and
then
a
birth
son
of
the
conscience
and
then
subtract
off
my
predicted
everywhere
from
that
data
set
and
see.
If
my
en
den
from
that
problem
matches
that
that
would
be
that's
the
ultimate
goal.