►
From YouTube: SimPEG meeting June 25 part 2
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
A
So
we
thought
I've
also
been
looking
at
remaining
modules
to
try
and
be
a
little
more
pythonic,
so
looking
at
using
lowercase
names
for
modules
and
also
trying
to
use
less
shorthand.
So,
for
example,
in
this
survey,
instead
of
s
are
seeing
it
list
I've
written
that
out
as
a
source
list.
It
will
still
let
you
use
SRC
list
is
just
gonna,
throw
a
warning,
and
eventually
we
can
start
to
delete
these.
A
So
this
is
what
I'm
saying
by
like
many
of
the
changes
we
can
make
somewhat
backwards,
compatible
to
try
and
not
have
this
be
too
painful
of
the
transition.
But
some
some
things
like
the
deep
red
is
non
backwards.
Compatible
I
could
throw
a
warning
or
I
can
throw
a
sensible
error,
but
there's
nothing.
We
can
do
to
like
make
the
survey
hand
deep
right
back.
It
won't
know
how
to
do
that
so
with
that
there's
also
I've
gone
in
and
suggested
some
naming
of
changes
to
module.
A
So,
for
example,
instead
of
saying
F
Tem
do
we
want
to
call
it
frequency
domain
so
from
simple
electromagnetics
import
frequency
domain
and
you
could
import
it
as
f
de
m
if
you
wanted
to
continue
like
having
sort
of
short
of
hands
in
your
Python
scripts
or
in
your
notebooks,
but
looking
at
favoring,
something
that's
a
little
more
explicit
I
think
it's
a
bit
more
newcomer
friendly,
but
that's
night,
I
realized
that
this
is
a
major
change
and
want
to
get
input
from
everybody
and
give
everyone
a
chance
to
chime
in
before
we
start
making
all
of
these
changes.
D
A
Now
the
frequency
domain
is
up
running
so
I'm
I've
migrated
MOOCs
to
the
e/m
codes
and
in
progress
on
the
DC
side
of
things.
So
the
static
codes
are
not
gonna
work.
Quite
yet.
I
think
DC
ship,
but
I
haven't
had
a
chance
to
go
through
DC
IP
and
s
IP
in
as
much
detail.
Yet.
Okay,
so
working
looking
on
that.
A
But
if
people
want
to
start
playing,
especially
with
frequency
domain
time
domain
or-
and
please
do
then,
hopefully
by
the
end
of
the
next
week
or
two
we'll
have
sort
of
first
pass
examples
of
them
all
up
in
Arnie.
But
if
you
have
it
has
to
chime
in
on
this
naming
thing.
I
would
really
appreciate
that,
because
I
can
start
doing
that
now,
but
it
it
is
gonna
be
like
a
a
bit
to
untangle
and
so
prefer
to
only
do
it
once
so.
Hopefully,
we
can
get
it
right.
A
The
first
time
so
yeah
feel
free
to
like
start
planning,
and
you
spend
all
your
thoughts.
I
tried,
Indian
q
or
in
the
comment
thread
to
outline
the
major
changes.
So
let
me
see
if
I
can
pull
it
up
and
share
my
screen,
real,
quick,
so
I
don't
know
if
you
haven't
seen.
A
Can
y'all
see
that
yeah,
okay,
yeah
so
tried
to
outline
just
like
high-level
major
changes?
I'll
keep
updating
this
as
I
go
but,
for
example,
the
survey
no
longer
has
a
deep
red
method
today.
Also
no
longer
has
an
eval
method,
I've
gone
in
and
tried
to
simplify
the
data
class
a
bit,
and
so,
if
you
want
to
jump
in
and
have
a
look
at
that,
that
would
be
great.
The
data
class
now
has
so
before.
It
was
very
unclear
where
you
were
actually
supposed
to
set
the
standard
deviation
in
the
noise
floor.
A
A
That
said,
if
you
wanted
to
go
in
and
mess
with,
like
the
phi
d,
the
w
matrix,
you
can
do
that,
but,
like
all
of
the
noise
information
is
going
to
be
set
on
the
data
class
yeah,
so
that
I
think,
is
basically
so
there's
the
main
comments
here
and
then
further
down.
I
have
left
a
few
questions,
but
the
main
things
are
looking.
Basically,
we've
got
a
conversation
thread
here,
going
niels
jumped
in
and
dieter's
jumped
in
so
feel
free
to
to
add
your
thoughts
yeah
and
we
can
iterate
from
there.
A
E
I
was
written
in
says,
I
think
that's
a
good
idea.
It's
like
here
are
words
to
understand,
but
are
you
also
going
to
omit
all
the
deputation
then?
Because
that's
important,
of
course,
is
that
superior
there
how
it
should
be
used
and
not
how
it
was
used
before,
and
that
looks
like
it's
a
lot
of
work.
C
A
Absolutely
that's
a
great
point
and
I
think
what
I
would
perhaps
like
to
do
with
this
pull
request,
because
it
is
so
major
is
basically
we
can
get
it
up
to
a
state
where
we're
we're
happy
with
it
and
consider
it
functioning
and
consider
things
up
to
date,
and
then
I
would
like
to
like
request
basically
that
everyone,
at
least
to
sort
of,
is
much
more
familiar
with
simple
start.
Relying
on
it.
A
I'm
not
yes,
I
will
send
out
messages
when,
when
we
think
it's
basically
ready
but
start
relying
on
it
and
like
just
jump
in
and
help
bug,
hunt
because
I
mean
we
have
very,
which
is
great.
It
makes
a
refactor
like
this
possible,
but
at
the
same
time,
like
they're,
still
always
little
things
that
slips
through
the
cracks
and
edge
cases
that
weren't
tested,
and
so
it's
going
to
take
yeah
everyone's
helped
to
make
sure
that
we
we
catch
all
that
stuff
and
then
also
like.
A
If
you
know
everybody
can
pick
a
different
aspect
of
the
examples
to
jump
in
and
help
with,
then
that
gives
only
each
person,
maybe
two
or
three
examples
to
look
at
and
to
make
sure
that
it's
up
to
date.
So
it
absolutely
is
a
lot
of
work
but
I
think
between
all
of
us.
We
can,
we
can
make
it
all
happen.
B
A
Basically,
basically,
I
don't
think
we
well.
We
can
try
to
both
like
Linux
of
how
to
actually
do
that,
but
basically
treating
it
in
a
few
weeks
time,
perhaps
like
a
look
at
there's
brain
that
will
build
off
of
from
there
and
to
this
like
preferences,
that
we
want
to
sort
of
merge
this
stuff
into
I
mean
either
clean
or
dry.
You
like
set
up
calls
to.
A
C
And
just
to
make
sure
I
understand,
sort
of
how
this
will
work
didn't
running
simulations
in
practice.
We're
basically
we're
gonna,
have
a
set
of
problems
and
specific
simulation
classes.
So,
whereas
we
plan
something
like
I,
do
we
have
one
global
simulation
class
where
we
be
put
in
just
like
a
survey
and
then
a
model,
and
it's
supposed
to
figure
it
out
or
we'll
be
created
like
we
just
like
normal
will
be
creating
like
I,
don't
know,
Jaime
electromagnetic
problems
simulation
and
that
will
accept
a
like
an
M
survey
object.
A
So
we'll
still
have
weight
from
electromagnetic
from
simple,
like
frequency
domain,
electro,
magnetics,
import
simulation,
3d,
J,
and
so
that
will
still
function
basically
very
similar
to
the
problem
that
we
had
before
the
problem
class
that
we
had
before.
But
you
no
longer
need
to
prepare.
You
know
how
we
had
that
network.
It's
like
probably
serving.
We
won't
do
that
anymore.
What
you'll
do
is
set
simulation,
dot
survey
equals
survey
or
you
can
pass
the
survey
when
you
instantiate
the
spinning
elation.
A
G
Lindsey
I
think
that'd
be
great
and
I
think
it
where
I
am
and
I'm.
You
know,
there's
there's
a
lot
of
things
that
seem
really
reasonable,
but
I'm
not
exactly
sure
I've
got.
You
know
a
complete
overview
of
the
whole
package,
though
they
might
be
able
to
contribute
in
some
way,
so
I
I,
think
sort
of
laying
it
out
like
okay
guys.
This
is
this
is
basically
what
we
are.
G
This
is
what
we're
doing
and
does
it
all
seemed
to
make
sense,
because
everybody
has
got
you're,
bringing
a
spectrum
of
problems
that
they've
been
interacting
with
over
the
years.
So
they've
got,
you
know,
survey
geometry
and
data
and
instruments.
You
know
just
want
to
make
sure
that
it
all
kind
of
fits
in
a
natural
way.
Yeah.
A
For
sure
for
sure,
and
also
prompted
some
thoughts
with
respect
to
how
we
want
to
follow,
driver
or
codes
that
brought
impact,
for
example-
and
this
is
a
conversation
we
can
continue
later
on,
but
I
really
like
what
plot
lead
has
done,
so
they
have
plotly,
which
is
the
core
library.
It's
a
really
low
level
API,
and
by
that
I
mean.
If
you
want
to
make
a
plot
with
core
plot,
then
what
they've
done
is
they've
built,
partly
expressed,
which
is
like
a
lightweight
linker
on
top
of
plotly.
A
That
is
like
makes
more
assumptions
for
you,
but
it
also
like
is
a
little
easier
to
use.
So
I
think
that
some
of
the
stuff
that
perhaps
has
been
done
by
by
soggy
and
Mike
and
everyone
in
the
IO
DC
class
and
then
by
Dom
in
the
driver
class.
You
could
start
thinking
about
how
to
have
that,
perhaps
in
like
a
separate
module
with
instant
PayPal
it--wait
wrapper.
So
that's
just
something
in
like
throw
out
there
and
we
can
think
about.
A
C
I've
got
tutorials
on
models
and
mapping.
Sara's.
Give
me
a
review
on
that
and
seems,
like
you
were
quite
happy
with
noted
state
of
those
tutorials
I,
think
you
won't
pass
the
Travis
test
due
to
some
hot
related
stuff,
I.
Think
Auto
generation
of
the
API,
but
yeah
I'd
like
to
merge
that
in
and
have
it
sit
there
for
a
long
time
so
undergone
a
review
process
and
I
think
it's
pretty
close
to
being
brought
in
so
I'd
like
to
do
it
before
we
forget
about
it
and
move
on
to
this
stuff.
A
Okay,
I
will
do
my
best
to
get
to
it.
There's
a
big
backlog
at
the
komyatti
right
now,
still
oh
yeah,
and
if
there's
other
people
willing
to
jump
in
and
help
I
know
Derek's
giving
it
over.
Can
you
yeah
I
mean
if
people
were
sonken
you're
dawn,
if
you're
also
willing
to
jump
in
how
be
great,
but
if
it's
not
we'll
we'll,
try
and
push
it
push
it
forward
and
feel
pretty
just
like
keep
hoping
this
weapon,
even
if
we
haven't
done,
if
we
haven't
done
anything
by
then
keep
bringing
it
up.
Okay,.