►
From YouTube: SimPEG Meeting Jan 9, 2018
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
C
A
A
A
F
C
A
C
A
G
A
D
D
B
So
maybe
yeah,
but
the
like
he's,
mostly
using
a
backward
Euler,
so
I
I.
D
B
D
A
C
A
A
C
A
So
four
things
that
I
want
to
pro
today
is
just
like
briefly:
we
got
the
two
disparate
eyes
cleric
Weston.
Now
we
can
go
over
those
briefly.
Thank
you
guys
and
then
the
TTIP
stuff
is
still
around
and
then
in.
A
Know
it
took
a
long
time
to
get
these
when
things
as
a
thank
you
guys
for
your
patience,
but
we
did
get.
There
were
two
releases
that
went
up
this
weekend
for
district
IDs,
because
now,
what
we're
doing
since
we're
merging
onto
master
every
merge
onto
master
is
a
new
release.
So
that's
also
why
it
doesn't
like
automatically
let
you
merge.
It
is
because
one
of
us
has
to
go
in
and
actually
like
operation
make
sure
the
release
is
gonna
go
through
so
so.
E
B
A
So
it
that
should
all
happen
automatically,
but
I
don't
know.
Well
now
it
is
there's
a
couple
things:
pi,
PI
or
KPI
is
they're
upgrading
their
code
weights
as
well.
So
there's
some
things
that
are
like
changing,
which
I
haven't
liked
them
too.
So
I'll
show
you
just
briefly
like
what
I
do
with
pull
requests,
and
this
also
just
then
plays
into
like
how
you
Braille,
how
you
write
a
pull
request,
sort
of
knowing
that
it's
going
into
the
release,
notes
so
I'll
always
just
highlight
which
PR
it's
from
they'll
be
a
title.
A
So
hopefully
it's
something
difficult,
ending
feature
or
whatever
I'll
highlight
who
committed
and
who
review
and
then
it'll
be
a
bit
of
a
brief
description.
So
I
come
from
you
guys
if
you
have
a
picture
of
something
to
include.
That
is
also
awesome.
It's
super
helpful
and
just
like
gives
context
to
people
if
there's
issues
that
you're
closing
or
that
are
related
to
the
floor
quest,
please
just
reference
those
in
there
just
like
this
comes
from
here.
Whatever
the
other
thing.
A
A
D
F
F
A
This
is
just
a
quick,
very
related
to
flow
requests,
some
thoughts
just
on
branch
naming
conventions.
This
is
helpful
because
often
the
branch
ends
up
being
the
title
for
the
floor
request
and
also
when
it's
in
progress.
It's
just
nice
for
people
to
have
a
bit
of
an
idea
of
what
you're
up
to
so
I
thought
we
might
try
and
follow
a
bit
of
a
convention.
A
So
here,
basically
it's
the
majority
of
the
changes
that
you're
making
your
life
regretting
anything
things
feats
would
be
a
new
feature,
so
you're
adding
bug
if
there
is
a
thing
that
is
broken
dogs
or
dogs
testing,
if
you're
adding
tests
yeah
example,
if
there's
a
new
example
and
then
TANF,
if
you
just
heard
like
and
intended
to
later
that
week,
this
branch-
that's
fine,
cuz,
it's
important
out
stuff
like
that.
A
A
A
C
E
E
A
D
E
A
That's
fine
Machinery,
whirring
yeah,
that's
awesome!
Okay,
so
this
one
is
like
ready
for
review
cool,
because
I
would
like
to
get
that
one
in
first
and
then
we
can
get
it
nice,
3d.
So
much
so
one
of
the
changes
I
want
to
bring
in
with
this
guy
is
also
that
the.
A
A
A
D
E
C
D
A
C
A
D
A
B
A
A
F
E
A
Need
to
have
a
look
again
I've
like
lost
track
of
this,
because
there's
so
much
to
change,
but
I
can
take
another
look
like
it's,
because
I
stopped,
because
people
were
pushing
a
watch
code
and
then
to
start
failing
so
I
just
dropped
it
and
figure
to
pick
it
up
again
once
it
works.
So
if
it's
working
then
I'll
take
a
look
and
then,
if
you
could
just
comment
on
where
you're
trying
to
mesh
IO
stuff
or
the
IO
stuff,
I'll
take
a
look
at
like
where
you're
going
with
that
and
see.
B
G
F
A
F
A
I
mean
I,
guess
like
looking
through
that
the
high
impact
points
to
like.
What's?
Are
you
saying
it's
the
I
of
stuff,
which
we
know
we're
gonna
change
it
like?
Don't
worry
about
it
if
it
is
actually
sort
of
eating
some
of
our
code
like
the
core
code,
but
you
want
to
be
thinking
it's
worth.
Having
a
look
at.
Oh
it's
more
much
like
it's,
not
something.
C
A
E
A
A
A
B
B
B
A
B
Sensitivity
and
the
next
video
is
much
higher
resolution.
I
like
its
space
I'm,
not
exactly
sure
why,
but
I
can
tell
like
yeah.
You
got
more
time
channel
that
actually
forces
you
to
get
there
yeah
so
that
there's
less
uniqueness,
probably
I,
don't
know
that
any
way
you
could.
You
could
I
think
that's
actually
hard
hard
example
right,
but
similar
body
attach
it
right.
Then
you
really
like
just
we
can't
do
that.
It's
very
hard
to
unravel
yeah,
but
by
doing
inversion,
inversion
connection
Ravel
those
two
different.
A
A
B
Let's
say
you
got
some
functions
that
that
you
can
generate
3d
survey.
Let's
say
you
got
up,
we
put
a
pole,
pole,
50
meter
and
50
meter
spacing
and
spacings
15
start
from,
like
always
in
line,
so
the
X
is
always
in
line.
You
got
your
pole,
hole
at
whole
life,
there's
always
next
direction,
and
you
just
put
it
most
aligns
it's.
B
Survey
but
that
superposition
of
to
the
survey-
and
this
is
a
function
like
so
I
grabbed
what
da
meant
evil
wrote
and
just
can't
whipping
over
that
yeah
and
output
of
survey
and
IO.
So
that's
and
then
you
can
actually
set
the
mesh
here.
So
it's
kind
of
well
other
things
are
default,
so
you
just
need
to
set
that
the
that
you
don't
need
to
like
there's
something
for
like
a
whole
page
hand.
Then
you
find
the
minimum
distance
thing.
Yeah,
I'm
gonna
put
at
least
four
cells
yeah.
B
Default,
but
that
sometimes
it's
too
much
so
you
can
just
be
by
yourself,
but
it's
simple
like
wait,
just
a
its
generate
mesh
and
active
cells,
she
have
its
poverty
and
we
can
extract
the
core
mesh.
It's
cuz
I.
Are
you
doing
that
from
your
survey?
You
can
kind
of
say:
okay,
this
is
my
Quartermain
right
yeah.
We
got
a
electrolocation
and
you
could
use
whatever
maximum
space
in
one
of
the
three
that
you
can
be
beside
and
maximum
depth
that
you
would
get
yeah
so
then
we
can
actually
decide.
Okay.
B
A
B
B
C
B
D
C
B
B
E
B
G
B
B
A
A
A
C
G
C
B
A
At
least
having
sort
of
like
a
standard
approach
to
the
yes
and
flavor
and
then
to
making
sure,
because
the
thing
that
I
guess
like
so
there's
benefits
and
drawbacks
to
making
it.
You
see
like
it's
great
when
you
can
remove
those
choices
that
you've
already
made
a
hundred
times.
But
in
this
case
like
you,
could
see
this
in
the
long
run
being
dangerous,
because
people.
A
They're
using
for
regularization
doses,
so
just
trying
to
strike
that
balance
of
transparency
and
ease
of
use,
yeah
but
I
think
there's
there's
ways
to
do
that
and
so
yeah.
So
yeah,
yes,
and
this.
A
A
E
C
A
And
then,
what's
nice
like,
if
we
think
this
through
properly
is
like
we'll
be
able
to
serialize
and
all
that
sort
of
and
then
like
automatically
generate
scripts
and
things
like
that,
because
that's
actually
one
thing
that
I
really
liked
it
with
the
case
inside
of
things
is
I
can
like
set
it
up
interactively
and
then
I
just
say,
write
my
script,
one
minute
simulation
and
then
I
kick
it
over
the
cluster.
That's
actually
because.
C
C
E
E
G
You
know
defend
other
things,
so
I've
made
a
whole
bunch
of
tests
and
I
have
some
comments
on
my
floor
West,
but
at
some
point
it
would
be
kind
of
nice,
and
so
you
can
sit
there.
I'll
be
just
itemized
all
of
the
remaining
things
to
get
that
whole
or
close
to
it.
So
they're
a
so
when
you're
talking
about
which
there
they
are
okay,
so
it
works.
G
If
it's
super
well
tested,
so
you
know,
does
the
American
solution
as
much
analytic
solution
was
trying
to
have
as
many
parts
of
it
who
isn't
and
but
I
think
I'm
not
touching,
and
you
go
up,
you
see
where
it
went
so
sort
of
internally
you're
on
its
own.
It
seems
like
I've
I've
touched
most
of
the
aspects
of
my
coat,
but
I
haven't
really
linked
it
or
got
it
to
interact
I.
G
G
A
A
So
that
then
just
needs
to
be
added
to
Travis,
so
I
can
point
you
to
where
that
goes,
and
then
the
property
side
of
things
I
can
try
and
point
you
to
a
few
things.
I
wouldn't
worry
about
it.
Much
yet
I
might
put
you
to
a
couple
high
priority
items
but
like
we
don't
have
to
fully
inject
it
into
yet
yeah.
G
B
G
A
A
G
A
G
A
G
G
G
G
Is
you
stuff,
then
yeah
for
now,
but
yeah,
and
then
the
other
thing
was
just
and
maybe
something
that
I
end
up
doing
I
really
like
to
see
on
the
big
website
and
probably
district
lines
as
well,
just
this
linear
sort
of
itemized
list
that
starts
with
so
you
want
to
enter
it.
You
want
to
make
a
branch
put
some
stuff
in
and
merge
it
in,
and
then
here's
our
stats
that
are
awesome
so
yeah.
What
do
I
need
my
tranche,
you
know,
writing
a
test
and
that
really
needs
to
get
absolutely
our
dog.
Sir.
G
A
It
wasn't
the
things
that
we
could
think
about
doing
is
pick
a
day,
maybe
and
there,
maybe,
when
you're
back
and
do
like
a
doc
upon
thanks,
it's
not
a
bad
way
to
get
stuff
done
right,
just
throw
a
day
at
it.
It's
better
that
way
to
get
a
squish,
yeah
and
because
I
know
that,
like
none
of
us,
it's
not
something
that
hits
the
radar,
because
it
is
sort
of
like
actual
work.
But
at
the
same
time
it's
like
this
tsunami
can.
G
A
Okay,
so
we
can
because,
but
I
think
like
what
I
would
like
to
have
done.
First,
is
that
we
just
sketch
out
a
template,
and
so
I
was
gonna,
try
and
me
something
from
the
e/m
side
of
things,
because
we've
got
so
much
content
like
we've
got
the
paper,
but
all
the
stuff
we
did
to
the
disk
like
it
should
come
together
into
something
from
here.
A
A
One
other
thing
I'm
chatting
with
Teeter
on
Thursday:
dear
by
tomorrow,
he
wrote
p.m.
hi
mom,
which
we
used
for
the
marine
CSM
example.
So
it's
a
1d,
but
you
can
do
an
X
and
I
set
repeat:
20
M
code
can
do
anisotropy.
Well,
that's
and
he
wants
to
plug
in
so
he's
got
the
Ford
simulation
analytical,
which
would
be
great
to
actually
have
hooked
up
to
sim
peg
a
little
more
closely.
But
then
he
wants
to
use
simple
for
the
inversion
side
of
things.
A
So
we're
gonna
start
chatting
about
how
to
interface
switches
or
shy.
Thing
will
be
kids,
because
I
think
we're
gonna
need
to
rethink
a
lot
of
how
we
do
the
Ford
problem
and
how
we
interface
to
it,
because
it's
where
your
new
cumbersome
and
it's
gonna
be
some
stuff
I
think
that's
gonna
come
to
light
once
we
start
diving
into
the
impro
that
we
need
to
simplify
because
right
now
we
don't
have
like
a
so.
F
A
F
F
F
C
It's
just
because
when
people
use
the
simplest
case,
rigorous
Asian
is
it's
just
a
list
of
objectives
right,
yeah,
but
it
more
complicated
cases.
This
is
also
going
to
be
enlisted
organization,
so
I
think
we're
agree
that
at
the
end
would
be
nice.
Maybe
the
different
problem
transforms
everything
into
compo,
objective
functions
and.
A
C
A
Dislike
in
a
lot
of
ways,
the
way
that
conceptually
I
sort
of
want
to
think
that
the
directive
is
is
actually
tied
to
like
a
piece
of
an
objective
function.
So,
for
example,
beta
beta
should
just
be
a
directive
that
I
can
just
say:
beta
equals
pudding,
schedule
whatever,
and
then
now
that
is
just
then
like
lumped
in
with
my.
A
A
A
A
C
A
C
A
A
The
other
thing
that
I
think
would
be
kind
of
cool
is
because,
right
now
the
iteration
printers
are
again
outside,
whereas
the
objective
function,
who
knows
what
it
has
so
it
should
know
how
to
look
at
print
itself
or
the
log
itself,
and
so
like
that
eventually
could
be
really
helpful.
It's
just
like
when
I'm
logging
I
would
love
to
see
this.
Isn't
this
funny
that
actually
should
not
eat
you
part
of
the
change
that
we
could
probably
do
to
mention.