►
From YouTube: SimPEG meeting February 5, 2019
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).
D
B
C
We
just
start
sure
so
the
meeting
notes
I
circulated
so
if
someone's
game
to
just
sort
of
like
fill
some
stuff
in
as
we
have
a
conversation,
that'd
be
great
but
yeah
thanks
Tom
as
Tebow
suggested.
Why
don't
we
go
around
the
room
and
sing
along.
B
Okay,
let
me
make
just
just
a
few
introductory
comments
here
for
everybody,
so
we're
very
privileged
to
have
three
people
who
are
going
to
join
us
for
six
or
seven
months,
there's
Sarah
Gauri,
who
is
live
and
well
on
the
on
the
screen.
So
she'll
tell
you
about
herself,
but
she
has
just
arrived
a
couple
of
days
ago
with
her
husband
and
two
lovely
young
children
and
they're.
Just
sort
of
starting
to
settle
in
they've
got
a
place
in
Vancouver
which
is
not
easily
combined,
and
so
there
yeah
she's
coming
into
voting.
B
Then
there's
Kevin
Pham,
who
has
just
joined
us
yesterday.
So
Kevin
was
a
university
graduate
here
and
has
been
doing
some
work
in
in
my
and
Maura
he's
now
going
to
be
coming
back
and
he's
going
to
work
with
us
or
throughout
the
summer
and
ariston
II,
who
is
now
a
new
Iranian
after
Mesa,
who
was
here
for
four
years,
turns
out
mace
salmon.
B
E
F
Teach
your
physics
in
Brazil
working
with
the
national
geophysical,
but
mainly
we
process,
analyzing
data.
The
last
recently
I
haven't
been
working,
especially
with
phenomenological
issues.
So
how
can
I
say:
Sun
the
earth,
so
the
geomagnetic
data
mainly
looking
for
positions
for
recurrent
recurrent
passing
geomagnetic
pole
in
the
past,
from
a
low
magnetic
data,
and
things
like
that
so
but
I
did
a
postdoc
in
in
real
Janeiro
in
national
observatory
in
Brazil,
working
to
my
hem,
some
electromagnetic
device,
low
integer
number
and
magnetic
lorica
also
and
right
now,
I
was
trying
to
to
meet.
G
B
And
one
other
thing
so
Dirk
also
teaches
an
introductory
course
geophysics
for
geologists
at
the
sea.
Okay,
so
he
is
really
the
only
I
guess
it's
really
the
only
geophysics
course
that's
available
to
them,
and
so
it's
very
very
similar
to
yo
the
your
on
my
note
for
that
that
we
have
and
right
now
for
a
textbook
he's
using
Reynolds
that
book,
but
he's
just
sort
of
talking
a
little
bit
through
the
you
know
the
GP,
GE
and
so
yeah.
B
He
might
expect
missing
parts
of
that,
except
there's
a
number
of
connections
both
with
respect
to
introduction
of
gift
for
geologists,
as
well
as
just
general
computational
geophysics,
because
George
is
very
interested
in
you
know
signal
processing
in
particular,
but
with
applications
to
electromagnetics
and
in
particular
you
know,
magnet
of
Tillery
stations
under
the
electric
jet
and
so
there's
yeah.
Quite
a
few,
quite
good
potential
connections
plucks.
He
started
out
as
a
Python
programmer
switched
in
MATLAB,
but
now
is.
C
D
That's
fine,
okay,
so
yeah
I'm,
Sarah
and
I'm
from
Belgium
I
am
a
professor
at
the
University
of
liège
on
the
bioengineering
campus
in
John
brew.
So
that
was
actually
for
150
years.
An
independent
Agricultural
University
and
is
now
adhere
to
the
University
of
liège.
There.
I'm
teaching,
not
really
I,
have
one
course
introductory
course
on
geophysics,
but
I'm
more
in
everything
that
is
related
to
soil
and
plant
interactions.
D
That's
also
one
of
the
reasons
I
contacted
you
to
come
here.
I
do
use
your
online
your
online
course
there
also.
So
what
I
have
my
own
slice,
but
then
often
I
tell
them
if
you
want
to
go
more
in
depth.
Check-Check
these
methods
online
in
your
in
your
textbook
or
how
should
I
call
it
your
information,
which.
D
Is
that
Manya?
Thank
you,
so
I
don't
use
it
like,
let's
say
actually
in
the
course,
but
I
refer
to
it
for
each
other
lesson,
I
said
the
links
online
because
I
think
it's
quite
convenient
for
them.
They
are
really.
They
don't
know
anything
about
it
and,
in
a
very
short
course,
they
get
a
lot
of
information
so
there
they
can
check
it
more
at
ease
because
I
did
have
Darrell
mass
blog
also,
but
I
I,
don't
like
that
one
very
much,
especially
not
for
ours,
because
that's
really
more,
like
your
we're
more
agricultural
scientists.
D
So
it's
the
difference.
Yeah
Reynolds
is
really
for
geologists,
I
have
the
impression
or
or
even
geophysics,
I,
don't
know
it's
another
scale,
so
I
liked
actually
more
what
you've
been
making
your
mind
anyway.
So
what
I
would
like
to
work
on
here
is
a
until
now
I've
been
using
here
tea,
mainly
at
small
scales,
on
in
lyza
meters
and
and
in
the
field.
That's
let's
say:
I
try
to
resolve
at
this
immediately
resolution.
D
G
G
From
s
to
talk,
chose
to
destroy,
University
actually
prefers
over,
there
are,
and
mainly
I
was
active
in
gravity
and
magnetic
as
a
lecturer
and
researcher
for
about
20
years.
I
did
many
projects
in
exploration,
geophysics,
specially
gravity
magnetic
and
so
I
have
a
good
experience
in
this
field.
I
knew
sir
software's
during
these
years
that
in
my
publication
and
my
interpretations,
for
example,
Camacho's
metal,
no
division,
the
genesis,
Joe
soft,
but
it's
my
first
time
that
I'm
becoming
familiar
with
this
graph
to
the
maitre
d.
This
is
the
main
reason
that
I
came
here.
G
Also
proposal
number
is
a
very
famous
calling
this
face,
but
it's
a
good
opportunity
for
me
to
be
here
for
my
sabbatical
period.
It's
about
nine
months,
one
maximum
one
year
and
I'm
very
happy
to
see
you
I
meet
you
here,
I'm
just
arrived
about
three
weeks
ago
and
now
I'm,
certainly
much
better.
It's
difficult
to
find
somewhere
to
relief.
F
G
I
Was
doing
housing,
affordability,
exact
tomorrow?
Ok,
let's
postpone
it
yeah,
so
I
I
did
my
geophysics
degree
30
BC
and
then
right
after
that,
I
went
to
Miami
to
do
a
six-month
volunteer
placement.
I.
Think
many
of
you
heard
this
story,
but
for
those
who
haven't
and
I
was
with
basically
I
was
volunteering
with
the
state-level
Department
in
southeastern
one.
My
am
mark
in
the
state
called
Mon
state
and
I
was
working.
I
was
working
inside
of
there
basically
to
train
the
engineers
in
geophysics.
I
Actually,
something
I
discovered
a
bit
lay
in
the
process
was
that
this
government
department
did
not
actually
have
any
geophysical
equipment
themselves.
So
you
know
this
is
a
really
big
problem.
How
can
I
teach
them
if
I
don't
I,
don't
have
any
equipment?
So
with
a
friend
of
mine,
we
found
this
designed
by
Professor
in
the
States
for
a
relatively
shallow
verdict:
electrical
sound
weakened
device
to
do
1d
service
and
we
built
that
device
with
some
funding
from
the
state
code,
a
state
government
in
Miami.
I
At
the
time
the
maximum
penetration
depth
was
only
about
30
or
35
meters.
So
for
relatively
shallow
shallow
investigations
and
basically
we
were
trying
to
train
them
in
how
to
state
wells
and
find
the
best
places
to
drill,
because
over
that
part
of
it
part
of
that
part
of
the
country.
There's
dry
season
lasting
for
about
two
months
and
basically
tens
of
thousands
of
people.
They
have
very
few
options
when
it
comes
to
water.
I
They
rely
on
company
donations,
but
if
the
government
doesn't
come,
then
there's
no
water
and
it's
really
tragic,
so
I
went
there
to
try
to
I
guess
help
build
capacity
with
the
government
and
I'd
say
we
got.
We
got
a
decent
start.
Basically,
we
were
able
to
test
to
calibrate
the
instrument
in
the
wheel
upgrade.
So
now
we
can
go
to
about
giving
90-meter
dubs
and
you
know
I
wish
I
had
more
time.
You
know
towards
the
end,
I
think.
I
If
I
had
a
few
more
months,
then
we'd
be
able
to
go
use
this
improve
device
and
do
better
surveys,
but
I
ran
out
of
time,
so
I
think,
with
with
gifs
and
with
everyone
here,
I'm
also
really
interested
in
open
source
materials.
We're
trying
to
get
funding
with
an
organization
called
geoscientists,
Without
Borders,
they're
part
of
that
secret
action,
Weaver
trying
to
get
funny
to
do
a
sort
of
follower
project
to
do
those
things.
I
B
A
Theoretical
physical
constraint
in
the
inversion
to
really
try
to
constrain
yeah
when
there's
a
recovered
result
and
like
a
force,
recovery
of
certain
key
features
that
we
want
to
see
so
I'm
trying
to
do
that
like
really
like
I'm,
not
like
I,
said
pretty
German,
but
let
the
algorithm
what
where
and
when
with
which
values.
So
we
we've
been.
We
we're
working
right
now
on
the
first
foundation
at
equal
with
drug
and
hopefully
by
the
end
of
the
week,
something
like
that
with
thee.
Okay,.
K
C
B
K
Haven't
really
so
that
the
iam,
the
extension
of
it
right
now
most
the
stuff
I've
been
doing,
is
on
the
DC,
resistivity
side
of
things
and
then
I'm
just
going
to
try
to
take
some
of
those
ideas
that
we've
developed
for
the
DC,
resistivity
and
kind
of
extend
them
into
the
electromagnetic
regime
and
kind
of
show
that
we
can
get
a
lot
more
information
from
the
electromagnetic
data.
Okay,.
D
My
questions,
I,
don't
know
whether
there's
some
specialist
here
around
the
table,
but
I
am
working
on
this
project
with
the
unbundle
cook
from
Munich.
You
probably
know
Anna
and
they
have
their
own
MATLAB
based
in
version
code
for
e/m
data.
But
I
was
wondering
in
how
far
we'd
sink
back.
We
could
do
a
1d
or
or
other
inversion
of
M
data
for
knots.
K
C
D
C
We
can
certainly
get
into
the
details
here
and
maybe
at
the
end
of
this,
we
can
try
and
compile
a
list
of
resources.
I
think
with
this
group
here
that
would
be
like
a
great
thing
to
figure
out
is
what
what
do
people
need,
and
what
are
you
looking
for
like
what
kind
of
questions
you
have
and
what
try
things
should
we
be
pointing
to
you
get
started,
okay,
what.
D
Was
the
one
thing
I
wouldn't
I
already
told
doc:
I
really
want
to
help
out
on
that,
because
I
think
probably
you'd
be
asking
a
lot
of
questions
to
you
guys,
and
maybe
what
I
can
give
back
during
my
stay
is
also
to
really
write
down
everything.
That's
for
me,
as
a
dummy
is
important
to
know
to
be
able
to
work
with
the
codes.
Maybe
you
can
make
some
kind
of
guide
for
dummies
than
that
with
all
these
things.
That
would
be
extremely
valuable.
D
B
J
Okay,
I
am
dong.
I
did
my
undergrad
here
at
UBC,
when
my
masters
were
drug
and
third
year,
I
guess
fourth
year
PhD.
So
if,
in
my
entire
academic
career,
mostly
focusing
on
potential
field,
potential
feels
like
my
graph.
F
did
also
help
us
work
with
the
same
peg,
dc-dc
R,
so
we'll
be
able
to
collaborate
a
little
bit
there.
On
that.
On
that
side,
you
know
my
focus
is
on
try
to
do
large
scale
right
now.
J
D
That's
really
good,
because
I
remember
when,
with
normal
transit
days,
I
don't
have
that
problem,
but
when
I
hit
my
smile
I
see
meter
data.
It
was
a
big
problem.
Actually,
today,
at
Birch
at
the
time,
was
not
paralyzed.
It's
used
to
take
like
I
know
a
day
to
to
invert
all
my
limiter
stuff
and
then
for
just
one
time
frame
so
actually
used
ten
nodes
at
the
same
time
on
the
cluster
to
be
able
to
do
yeah
a
long
time.
D
B
L
B
C
Like
so
yeah,
we
actually
started
at
the
same
time
and
finished
up
around
the
same
time.
So
I
did
my
PhD,
mostly
focused
on
electro
magnetics,
doing
a
lot
of
work
looking
at
Steelcase
wells
and
I've
done
a
lot
of
development
in
some
paid
all
throat,
as
well
as
with
like
the
GSI
resources
and
sort
of
open
educational
content.
Now
I
am
at
UC
Berkeley
and
I'm,
working
with
the
Jupiter
team,
and
so
a
lot
of
my
role
here
right
now
is
figuring
out.
C
B
B
C
There
are
a
few
items
on
the
list
for
there's
dogs
at
at
a
point
on
progress
hitting
on
tasks
for
the
combo
objective
functions.
So
those
of
you
who
don't
know
what
desk
is
task
is
a
great
way
to
parallelize
codes
in
python.
Don't
want
to
maybe
give
us
like
a
high
level,
and
then
we
can
sort
of
go
into
the
details.
Perhaps
next
week
appropriate
yeah.
J
That
sounds
okay,
okay,
so
for
those
the
new
new
person,
we
try
to
monetize
everything
insane.
Thank
you
right.
So,
okay,
think
of
the
math
we
have
like
the
regularization,
and
then
we
have
the
forward
problem
and
all
this
and
all
these
other
blocks
are
their
building
blocks
and
they
form
an
inversion,
something
like
more
complicated
them,
and
so
now
we're
trying
to
paralyze
like
we
paralyze
the
forward
problem
and
now
we're
trying
to
paralyze
the
inversion.
J
D
J
Does
both
yeah,
so
you
could
do
hopen
NP
style,
while
you
send
it
to
many
workers,
or
you
can
do
like
like
an
MPI,
sorry
MPI.
We
since
many
workers
or
OpenMP
a
Multi
multi
threaded,
so
it
does
both.
Okay,
that's
does
it
is
that
we
don't
need
multiple
libraries
who
can
do
it
all
just
with
the
swim
past.
C
J
That
I
think
the
only
issue
we
have
right
now
is
the
law
I
kind
of
figured
out,
but
the
smartest
consideration
with
tasks
is
a
bit
complicated.
Yeah.
It
needs
to
be
needs
to
be
treated
of
like
carefully,
because
every
time
you
end
up
with
the
sparse
matrix
operation
that
it
bottlenecks
there.
So
we
need
to
yeah,
but
yeah
run
some
tests.
Can't
you
guys
can
see
my
screen
here,
just
that,
it's
very
briefly
so,
usually
in
we
have
like
their
ford
classes
like
this
right.
So
we
have.
This
is
a
Python
class.
J
That
does
something
don't
know
you
don't
worry
about
it
here,
I'm,
just
creating
a
random
matrix
with
10,000
by
10,000
dimensions
and
then
the
four
bubble,
just
a
matrix
product,
actually
take
our
product
so
the
a
matrix
time
anytime
a
model
and
then
that
because
data
right,
it's
just
something
something
simple.
Okay,
so
in.
J
Sin
bag.
If
you
have
multiple
problems
that
CI
multiple,
like
transmitters
John,
you
were
just
doing
a
simple
loop
for
all
the
transmitters
and
doing
this
doing
this
dot
product
and
series
right.
So
maybe
this
heart
is
paralyzed,
but
the
loop
over
all
the
sources
is
not
that's
just
a
straight
up
for
you,
and
so
now
we're
thinking,
that's
just
paralyzes
sporty.because
yeah
do
all
the
for
models
are
the
same
time
if
you
have
multiple
processors
and
we
just
throw
a
lot
more
processors
of
the
problem,
and
this
is
what
ask
it
does.
J
That's
the
decorator.
You
can
just
ask
that
function
to
wait
and
then
that's
just
builds
up
all
the
all
the
functions
that
you
want
to
do
and
then,
when
you
ask
to
compute
it
creates
a
graph
and
then
send
it
to
all
the
processors-
and
this
is
the
runtime
in
series
versus
are
on
time
and
in
full
parallel
for
the
fourth
problem,
so
we'll
be
able
to
have
a
huge
speed
up
and
I'll
be
still
more
sources.
You
have
the
better
it
feels
like
this
is.
J
J
Alright,
so
that's
just
a
way
to
display
that
you
brought
the
processes
that
are
happening,
so
it's
the
same
problem
and
a
Jupiter
ANOVA
and
I'm
just
going
to
show
you
how
the
computers
are
interacting
with
us.
So
this
is
the
problem
in
series,
okay,
and
so
ever
time
you
start
the
Yolanda
process.
Today
it
recreates
a
new
graph
and
it
sends
a
check
out
on
that
on
the
processors
one
after
the
other
and
that's
what
the
part
that
takes
done.
J
D
F
D
J
Okay,
so
you
started
out
there
not
now,
if
we
do
the
whole
thing
in
parallel
and
said,
like
the
full
paralyzation
you'll
see,
the
graph
is
going
to
be
super
wide
and
that's
that's
what
you
want.
You
want
a
very
white
graph
and
then
you
want
all
the
process
of
UPC
all
the
time
right.
So
the
red
is
what
is
being
processed
and
green.
What
is
what
is
waiting
and
then
not
all
my
processors
are
just
working
full
time,
all
the
time
hundred
percent.
There
is
no
there's
no
downtime
between
each
four
models.
H
D
C
L
Kind
of
you
know
it's
like
a
two-level
paralyzation
problem.
Like
let's
say
you
got
a
simulation
level
like
you,
use,
10,
10,
processor,
or
something
like
that.
But
you
got
an
inversion
level.
Let's
say
like
so
you,
let's
say
you
got
total,
can
under
processor
and
I
kind
of
did
you
figure
it
out
in
like
a
higher
level?
And
then
you
have
also
another
level
that
that
can
you
know
what
I
mean
like
a
true
level
problem,
paralyzation
yeah.
J
So
the
highest
level
side
is
on
the
someone
when
I
played
fields
or
the
derivatives
right
on
the
convoy
objective
function.
So
currently
we're
looping
over
right.
If
you
put
the
decorator
at
that
level,
it
will
detect
if
there
is
any
pedestrianisation
downstream.
If
it
doesn't,
then
anyone
could
consider
them
as
a
as
a
as
a
series,
but
if
he
secretly
best,
then
they
will
add
them
to
the
graph
and
it
will
optimize
it
down
downstream.
It's
pretty
really
easy.
That's
doesn't
you
three
use
figures.
C
All
that
out
for
you
as
long
as
you're,
using
gasp
all
the
way
through
I,
think
if
you
start
mixing
multi-processing
and
gasps
I
need
to
specify
the
number
of
processors
on
multi-process
and
I
think
that
gets
messy.
But
you
used
ask
the
whole
way
through
dots,
just
figures
out
the
graph
for
you
and
then
you're
not
to
worry
about
it.
Yeah.
I
C
C
We've
created
a
bunch
but
they're
sort
of
scattered
all
over
the
place
at
the
moment,
and
so
this
has
been
on
the
radar
for
a
while
is
to
create
a
site
called
inversion,
dot
cos
I
dot,
X
Y
Z,
and
this
very
much
could
supplement
a
lot
of
the
simpe
documentation.
So
the
simpad
occupation
would,
sir
assume
that
you
already
understand
all
the
details
of
conversion
and
just
walk
you
through
like
how
to
piece
and
Ambersons
button
version
dot
geo
side.
C
That
XYZ
should
provide
like
an
overview
of
various
components
and
various
decisions
that
you're
making
along
the
way
like
our
designer
matched.
What
different
regularization
is
you
looking
at
things
like
the
beta
cooling
schedule,
stuff
like
what
we've
done
in
the
MT
tutorial?
What
Phil's
done
in
all
the
directed
studying
stuff
that
you
did
so
right
now?
I
think
the
biggest
thing
that
we
can
do
is
first
of
all,
assemble
everything.
That's
out
there
that
we've
done
so
I
started
with
a
brain
dump
of
what
things
are
that
I
can
remember
that
we've
done
so.
C
C
Everyone
is
like
what
are
the
questions
that
you're
trying
to
answer
and
how
like
what
are
the
things,
perhaps
that
simpe
you're
interested
in
learning
about,
and
you
would
be
interested
in
helping
rate
a
tutorial,
because
then
we
can
find
a
place
as
to
where
that
would
fit
point
you
to
the
right
resources
and
then
have
that
be
something
that's
captured
and
then
can
help
people
in
the
future
get
on
board
with
synthetic
and
understand.
What's
going
on
there
yeah
so
Doug
are
there
things
with
that
that
you
would
like
to
to
mention
as
well.
B
No
specifically
I
think
the
main
thing
here
is
is
for
us
to
yeah
just
get
a
place
that
we
can
start
to
put
all
of
the
material
that
we
want
to
have
for
current
version
and
then
to
gradually
build
up.
You
know
a
sequence
of
notebooks
that
allow
the
user
to
kind
of
get
through
all
of
the
really
important
parts
of
this.
B
So,
whether
it's
you
know
just
for
you
guys
about
just
all
the
elements
that
Lindsey
has
said,
and
we
do
have
a
lot
of
things
that
we've
kind
of
done
little
bits
and
pieces
here
and
there
so
I
think
this
is
an
opportunity
now
to
really
take
a
step
back
and
then
put
something
together.
That's
useful
not
only
for
the
current
people
here
now,
but
if
we
actually,
if
we
satisfy
everybody
who's
sitting
around
the
table
here,
that
this
is
really
all
I
need.
B
I
could
just
go
to
this
website
and
I'd
have
a
basically
a
self-directed
routine
here
for
how
to
get
through
how
to
understand
that
version.
How
to
implement
it
in
in
senpai
it'd
be
great,
and
that
would
be
a
big
checkmark
and
then
other
people
could
undoubtedly
handle
it.
So
that's
really
the
the
goal
and
I
think
you
know,
especially
over
the
next
month,
to
really
try
to
see
how
far
we
could
go
with
that
perspective,
developing
some
notebooks
and
really
making
it
everything
understandable
and
useful.
B
I
B
B
There
are
so
many
people
out
there
who
had
access
to
3d
and
version
codes
and
specially
magnet
alert
codes,
but
other
types
too,
and
you
know
they're
they're,
putting
up
stuff
and
putting
up
resolves
and
you're.
Looking
at
those
are
thought
you
said:
there's
there
is
no
way
the
earth
could
ever
ever.
Look
like
that
right,
and
so
you
start
to
talk
to
them
about
ok,
what
they
they've
got.
This
oh
wow
they've
got
this
3d
and
burning
code
great,
and
then
they
start
offering
the
right
words
like.
B
Oh,
they
start,
you
know,
here's
my
regularization
function,
here's
my
misfit,
here's
my
chicken
off
curve
right
so
big
old
words,
and
then
you
ask
them
something:
that's
just
a
little
bit
lower
down
and
it's
like
it's
not
there.
So
at
the
end
of
the
day,
they
end
up
basically
bringing
stuff
in
pushing
some
buttons
until
they
got
a
picture
at
the
outset
and
have
no
way
of
critically
thinking
yeah,
that
that
is
a
possible
good
answer
or-
and
so
it's
in
really
trying
to
emphasize
those
fundamentals,
emphasize
how
the
critical
thinking
for
inversion
works.
B
B
C
And
so
sorry
Sarah,
when
you're
bungee
jumping
no
sorry,
no
okay.
So
like
the
resources
that
we
put
together
and
click
the
g,
PG
and
E
M,
we
sort
of
don't
really
talk
much
about
code
and
don't
really
get
into
that.
This
could
be
different.
So
this
we
will
like
show
people
code.
How
do
you
use
a
mesh
on
like
how
to
build
a
map?
Things
are
how
like
kind
of
mulling
works.
All
of
that
for
themself,
so
I,
see
this
like
very
much
is
like
a
next
level.
C
C
C
If
you
people
need
to
just
like
flip
through
these
resources
and
then
also
I
started
in
the
notes
there,
just
a
section
on
questions
and
like
if
you
just
have
general
questions
about
emergent
or
general
questions
about
using
some
more
on
methods
at
this
point,
let's
sort
of
technical
like
how
do
I
know
run
in
joint
inversion
is
something
much
more
on
like
okay,
what
does
if
you
choose
an
l1
norm
versus
melts
new
norm?
What
is
that
une
inversion?
D
Sounds
good
yeah!
That's
no
problem!
Just
this
kind
of
questions
is
probably
I'm
already
a
bit
biased
because
yeah
I
used
them
already
for
a
few
years
and
it's
mainly
from
the
DC
resistivity
aspect
that
I'm
thinking
but
yeah
I
can
go
through
it
and
see
what
what
I
have
this
question
still
left
and
what
I
think
of
my
students
always
ask
me
and
things
like
that:
yeah.
C
D
D
C
D
Think
the
thing
with
norms
is
quite
important,
and
especially
whether
you
have
this
also
in
simpang
but,
for
example,
making
the
difference.
We
have
a
lot
of
things
with
time-lapse
and
so
to
explain
that
you
can
have
difference
regularization
in
different
dimensions
and
stuff
like
that
is
also
interesting
in
which
would
go
where
or
which
would
be
most
useful,
for
which
dimension
would
also
be
interesting.
For
example,
yeah.
C
Absolutely
so,
like
any
of
those
kinds
of
questions,
any
totally
feel
free
to
take
whatever
bias
you
can
channel
in
here.
This
is
make
it
so
where
should
they
go
to
to
put
these
cones
and
questions
in
the
Google
Doc
Google
dog,
stuff,
okay,
yeah?
We
can
keep
it
in
here
for
now
and
then
we
can
parse
it
into
something
in
the
communities.