►
From YouTube: SunPy Developer Discussion (18/4)
Description
A
Anyone
got
anything
they
will
want
to
all
go
out
as
we
get
started
or
wow.
This
is
slow.
What's
slow,
corlath.
A
A
A
A
D
C
A
B
B
B
Know
that
the
the
name
so
the
one
thing
I
did
I
just
tried
right
now
is
the
same
thing.
That's
in
the
code
blocks
in
the
pr
I--.
I
have.
A
A
A
D
E
Sure
thing:
okay,
cool,
so
yeah.
Let
me
stop
sharing
screen
share
my.
E
E
E
E
So
the
primary
way
of
doing
that
at
the
moment,
because
I'll
see
the
corona
itself
is
too
thin
and
tenuous
view
to
get
freedom
into
the
structure.
You
just
get
a
levitating
rope,
I
think
you're,
looking
up
see
directly
at
the
Sun
youth,
unfortunately
can't
see
the
contrast
to
the
corona
versus
the
solar
surface.
Art
is
even
semen-demon
effectively
ban
on
that
take
fat,
but
you
can't
do
on
the
corona.
E
So
we
tend
to
take
the
magnetic
gram
of
the
surface
and
extrapolate
above
that
using
various
extrapolation
routines,
the
simplest
of
which,
being
the
potential
extrapolator
and
more
complex
ones,
generally
being
nonlinear
force,
free
field
models,
and
so
my
project
was
essentially
to
create
an
extrapolation
routine
which,
in
the
end,
I
did
the
potential
extrapolator,
because
it
was
only
a
dissertation
project
and
I'm
also
trying
to
evaluate
that.
So
I
I
found
the
title
21
in
equilibrium,
analytical
solution,
which
is
meant
to
be
a
an
analytical
model
for
roughly
the
sort
of
magnetic
field.
E
You'd
expect
to
find
around
an
active
region,
including
the
the
flux,
tube
and
obviously
the
footprint
of
the
flux,
tube
and
then
using
that
to
evaluate
the
field.
So
what
my
project
for
the
Google's,
so
the
ISA
summer
of
code
in
space
was,
is
trying
to
take
the
code
I
generated
for
my
project
and
turning
that
into
an
api.
E
So
one
people
have
access
to
the
actual
code
that
I
created,
but
to
also
people
access
to
an
API
which
allows
them
to
put
in
their
own
extrapolation
routines
and
use
the
tools
that
I
provide
them
to
potentially
evaluate
those
those
fields.
So,
among
other
things,
one
of
the
sort
of
major
things
I
created
was
I
created.
The
3d
object
class,
which
is
the
3d
map
class
map
3d
class,
which
is
essentially
similar
to
a
normal
map
in
2d
stores.
E
The
data
as
a
son
as
I
as
a
numpy
ray,
but
obviously
it's
a
Eva,
3d
or
4d,
depending
whether
you've
got
a
scalar
field
or
a
vector
field,
numpy
array
and
it
has
all
of
the
normal
functionality,
that's
associated
with
a
map
you're
able
to
resample
you're
able
to
obviously
yeah
so
I'm,
trying
to
think
crop
trying
to
take
the
word
crop
a
sub
map
and
you're
able
to
evaluate
the
data
within
their
so
yeah.
Essentially,
the
whole
idea
is
that
now
this
is
this
an
affiliated
package.
E
That's
not
part
of
the
main
son
pie
talk
set,
but
assuming
you
install
my
package
you're
now
able
to
start
doing
extrapolation
based
analysis
all
very
simply
from
using
some
pie
style
input.
So
if
I
bring
up
right
so
be
the
entire
map,
3d
class
is
reasonably
big.
Can
people
actually
read
this
or
do
you
need
larger
font?
I.
F
Can
actually
see
I
just
got
a
black
screen
clip
for
you,
I
don't
have
anyone
elses
have
that
problem?
No,
I
can
actually
read
it,
which
is
surprising
to
me
of
y.
A
F
E
F
E
E
E
This
page
right
so
obviously
I've
been
making
rather
ridiculous
documentation
over
the
time,
so
everything
I've
created
has
hopefully
got
a
full
set
of
documentation
that
relates
to
it
map
3d
class
and
similar.
The
one
of
the
big
things
that
Stu
and
I
have
been
working
on,
and
it's
been
kind
of
learning
exercise.
E
If
I
sort
of
German
demonstrate
within
the
code,
what
you'd
need
to
do
in
order
to
do
that,
so
I
compare
extrapolates
generator
build
so
essentially
off
see
you
got
your
general
imports,
including,
if
you
want
to
visualize,
then
I've
got
a
number
of
visualization
fractions.
The
one
thing
that
stew
and
I've
been
working
on.
It's
trying
to
make
sure
they
are
ostensibly.
E
They
enable
you
to
access
the
my
of
I
interface
afterwards.
So
you
use
my
advice
for
plotting
because
that's
very
good
for
3d
plotting,
but
what
you
get
back.
You
get
back
access
to
all
of
the
objects
that
were
created
so
I'll
show
you
in
the
GUI,
but
essentially
means
that
you
should
be
able
to
then
modify
your
imagine
imagine
you
did
peek
on
a
sun
pie
map
in
the
same
way
that
you
potentially
can
access
parts
within
that
and
create
a
more
advanced
plot
from
it.
E
The
peak
is
just
the
very
start
of
it,
whereas
in
this
in
this
case
these
are
almost
the
functions.
I've
got
are
like
peak
and
then
from
that
you
can
modify
any
settings.
You
want
to
change
their
headings,
change
the
axes,
son
and
so
forth.
So
you
need
to
define
a
region
if
you're
trying
to
create
an
analytical
model,
you
obviously
need
to
define
a
region
in
physical
parameters.
As
with
some
pie,
I've
made
my
maps
unit
a
strip
I
unit
compliant.
E
There
has
been
a
slight
issue
with
the
fact
that,
obviously,
on
a
map,
you've
got
angular
units
where's
on
a
3d
volume.
You
obviously
can't
use
the
angular
units,
at
least
not
on
the
z-axis.
That
would
just
be
irrational.
It
wouldn't
make
sense,
so
there's
a
certain
amount
of
translation
that
may
be
necessary,
but
some
people,
if
they
want
to
start
converting
between
the
two,
but
essentially
you
just
set
up
a
friends.
E
It's
ethical
model
which
in
this
case,
is
the
title
in
your
equilibrium,
the
code
for
that
is
already
in
there.
You
then
want
to
generate
the
3d
map
based
on
that,
so
this
creates
a
3d
map
object.
Please
note
that
I
intend
to
sort
of
need,
not
the
naming
conventions
with
my
objects.
I
sort
of
fur
I've
been
a
little
bit
lacks
when
creating
the
examples,
just
because
I'm
still
trying
to
get
used
to
trying
to
create
more
professional-looking
code.
E
All
you
need
to
do
is
create
streamline
a
streamlined
object
using
the
streamline
plot
and
throw
it
that
3d
map,
which
is
generally
or
any
3d
maps
that
you've
created
slightly
more
complex.
If
you
want
to
add
the
boundary
data,
so
of
course,
you're
now
adding
a
2d
plot
to
your
3d
plot,
so
you
can
see
where
what
what
original
data
generate
that
then
there's
a
boundary
plot
function
and
then
similar
again
as
an
additional
specific
to
this
analytical
model.
E
You
can
add
the
TT
guidelines
and
those
just
gives
you
some
visual
cues,
which
are
shown
in
a
second
so
essentially
to
visualize
just
the
magnetic
field.
All
you
need
is
one
single
part
from
this
again,
you
get
all
of
your
outputs
which
enable
you
to
modify
it
exactly
the
same.
If
you
want
to
do
an
extrapolation,
all
you
need
to
do
is
create
a
potential
extrapolator
again.
E
It
needs
to
understand
the
x-ray
into
our
inch
and
said
range
with
the
volume
that
you
need
to
extrapolate
and
in
fact,
in
this
case
the
Z
shape
value.
So
obviously
how
many
increments
for
your
grid,
you
want
to
give
it.
It
uses
cuboid
grids
as
a
general
rule
in
theory
that
should
work.
If
you
had
a
polar
coordinate
system,
I
haven't
tried
that
like,
but
the
main
point
is
to
say
that
it
doesn't
have
to
be
cubic,
it
can
be
cuboid,
so
they
need
pixels
or
the
pixels
of
your
input.
E
So
if
I
run
this
fingers
crossed,
it
would
take
a
couple
of
seconds.
I'll
sit
actually
runs
model
and
its
various
it
over
in
my
of
a
window.
So
this
is
the
kind
of
plot
that
you
can
get,
so
it
needs
to
refresh
their.
So
this
is
just
a
streamline.
The
map.
Seed
object
is
just
a
normal
part
of
the
my
interface
you
can.
You
can
modify
that
if
you
want
that's
actually
a
widget
which
gives
you
a
10
by
10
grid.
You
can
modify
set
on
that.
E
Any
was
saying
that
the
whole
point
is
it's
meant
to
match
the
Maya
by
interfacing,
so
hopefully
yeah.
This
looks
reasonably
neat.
You
can
start
to
see
the
structure
of
the
Maya
by
scene,
so
vector
field
boundary
data,
and
then
these
are
all
just
guidelines,
so
I
could
actually
delete
these
to
make.
It
may
be
slightly
easier
to
view
so
that
that
was
the
flux
tube
which
I
got
rid
of,
which
relates
to
this
current
flowing,
and
so
you
can
actually
see
the
footprints
of
the
flux
tube
in
this
boundary
data.
E
There's,
no
reason
why
the
boundary
data
has
to
be
limited
to
this
cube.
You
could
create
the
ground
you
Gator
to
a
wider
axis
and
equally
off,
so
you
can
change
the
region
of
which
you
are
interested
in
simply
by
changing
the
parameters
you
put
within
the
model
or
by
resampling
afterwards
or
something
afterwards,
and
the
whole
idea
is
again
these.
These
are
just
cues
for
the
tightest
Molly
equally
pole
idea
is.
It
should
make
it
very
easy
to,
among
other
things,
visualize
magnetic
fields
again.
E
This
is
just
a
widget,
so
within
here
I've
got
streamlines
widget,
which
I
could
change
from
a
I
could
change
to
ribbons
if
I
wanted
or
tubes
all
of
that
sort
of
stuff.
This
is
all
part
of
the
Maya
by
interface.
The
logic
is
that
essentially
from
either
within
Python
or
using
the
the
GUI,
my
fi,
you
can
change
the
parameters
of
your
diagrams
and
there's
essentially
my
tools.
E
Just
let
you
get
easy
access
to
that
as
long
as
using
a
3d
map
object,
it
makes
it
very
simple
to
plot
these
things
and
equally,
the
boundary
map.
It's
worse
than
saying
the
boundary
map
is
a
completely
valid
valid
map
object,
so
I
can
plot
the
boundary
map.
So
this
is
the
XO.
This
is
the
analytical
solution
for
that
field,
remembering
that
the
potential
field
extrapolator,
is
a
very,
very
rough
and
dirty
way
to
get
a
magnetic
field.
E
E
E
E
E
And
and
just
do
this
I'll
show
it,
and
so
now
we've
we
simply
resampled
it
to
make
it
a
slightly
more
reasonable
field.
Obviously,
there's
a
lot
of
testing
when
it
comes
to
the
actual
models
and
making
sure
that's
correct.
I
have
actually
noticed
that
on
one
particular
analytical
model,
but
I've
somehow
missed
up
on
the
x
and
y
axis,
so
I
think
it's
only
on
the
boundary
data.
I
need
to
look
into
that.
So
that's
something
for
me
to
look
at
again.
E
G
E
Oh,
fine,
sorry
just
I
know
this
thing
has
been
a
bit
temperamental
for
us,
so
it's
just
moving,
but
everybody
else
is
Eva
stationary
and
I
got
movement
now
so
yeah.
The
main
thing
is
I've
tried
to
make
a
number
of
tools
which
enable
you
to
visualize
and
compare
so
in
fact,
there's
a
comparing
class
for
comparing
both
scalar
and
vector
fields
that
is
still
in
development.
So
at
the
moment
it's
quite
quite
limited,
but
hopefully
it
becomes
a
useful
tool.
E
I'm
hoping
that
I
I've
got
a
few
people
who
I've
spoken
to,
who
might
actually
be
interested
in
using
it
PhD
students
so
I'm,
not
necessarily
like
the
ultimate
professors
in
their
field.
But
at
least
it's
something.
I
don't
know
whether
there's
any
more
information
people
like
specifically
to
know
or
areas
they'd
like
me
to
show
you
them.
E
Yes
and
again,
there's
a
there's,
a
lot
of
examples
going
through
all
of
the
nitty-gritty
on
how
to
how
to
do
stuff,
for
instance,
how
to
just
do
an
extrapolation
of
a
natural
HMI,
and
if
it's
worth
me
saying
that
again,
I
did
mention
it,
but
watch
me
demonstrating
that
you
can
actually
write,
compare
extrapolating
generator
fields.
So,
in
theory
the
2d
map
should,
if
I've
made
an
error,
Oh.
E
E
A
A
E
E
We
are
better
so,
for
example,
that
was
the
fact
that
the
effective
magnetic
ram
so
lion
site
magnetic
ram
generated
by
the
analytical
model.
It
is
simply
a
fully
valid
map
that
can
be
analyzed
with
any
normal.
It's
unpaid
tool,
yeah.
So
that's
sort
of
what
I
wanted
to
show
and
see
again,
you
can
all
this
all
github
the
documentation
is,
there's
a
lot
of
it
there.
Obviously
it's
still
working
progress
up
with.
E
As
with
everything,
I
feel
that
at
least
it's
it
stands
on
its
own
now
as
a
functional
tool,
rather
than
a
few
months
ago,
it
was
still
kind
of
very
sketchy
or
three
areas.
Stuart
can
you
think
of
anything
worse,
be
demonstrating
from
that
point
onwards,.
E
J
I
C
E
J
A
C
E
You
remember,
I
was
at
the
time
I
started,
creating
it
was
specifically
for
magnetic
fields,
but
it
supports
scalar
and
vector
fields.
The
functionality
I've
tried
to
sort
of
mirror
what
functions
you
have
with
normal
maps.
Obviously
accepting
the
fact
that
it's
not
quite
a
normal
map.
For
instance,
the
data
is
stored
in
metadata
files,
a
metadata.
E
The
parameters
are
reasonably
similar,
ex-ranger
I
range
said
rains
rather
than
just
x
and
y
range.
Obviously
there's
the
issue
there
is
that
annoying
issue
between
when
you're
talking
about
to
the
observational
data,
which
is
again
measured
in
a
rather
complex
coordinate
system
versus
a
3d
dataset,
which
is
pretty
much
always
physically
physical
unit.
E
J
Maybe
how
about
we
hold
it
as
an
issue,
and
some
can
somebody
create
an
issue
that
says
you
know,
create
some
object
and
some
pie
that
first
that's
appropriate
for
3d
simulations,
and
then
we
can
have
some
discussions
about
what
we
need
and
what
we
want
there
and
you
can
refer
to
the
object
that
you've
developed
in
your
branch.
Just
just
so.
People
know
that
it
exists
that
we're
thinking
about
it
since
we're
not
apparently
going
to
move
forward
with
it,
putting
it
in
right
now
that.
E
J
E
E
J
A
A
A
I
was
bad
so,
which
is
that
it's
not
really
a
light
curve
anymore.
It's
more
generic
than
things
that
you
would
necessarily
consider
my
curve.
A
J
I
J
A
E
Like
I
would
come
up
with
that
kind
of
thing
off
the
top
of
my
head,
yeah
I
I,
haven't
used
like
like
her
that
much
it's
it
at
all
of
the
diagram.
No,
it
was
worth
me
also
mentioning
so
Stuart
and
I
had
the
conversation
about
a
week
ago
now
about
I
could
apply
for
the
east
of
summer
of
open
space
again
with
his
legs,
saying
that
I've
got
a
PhD
lined
up
in
Hull,
studying
the
possibilities
of
probabilities
of
superfluous
on
son.
E
E
I'm
doing
but
to
be
honest,
I'm
sort
of
feeling
that
I'm
already
spending
time
on
it
as
a
hobby
and
I
sort
of
feel
that
there's
not
a
lot
that
Issa
would
sort
of
be
getting
for
their
money
if
they
paid
do
it,
but
equally
finding
something
else
useful
to
do.
It
was
suggesting
potentially
refactoring
the
light
curve.
As
a
possibility,
so
I
don't
know
what
Bob's
other
people
think
are
important
and
are
unlikely
to
be
got
filled
in
by
google
Summer
of
Code
students.
E
J
A
J
A
Will
catch
you
up
when
we're
not
live
on
YouTube?
Okay,
because
otherwise
Google
get
very
angry,
very
angry
girl
and
be
Alex
I
mean
you
should
definitely
apply
it
and
you
should
probably
wait
and
see
what
projects
get
accepted
through
google
Summer
of
Code
first
and
then
haven't
liked
any
funds
still
around
and
come
up
with
another
100k
to
see
a
light
curve
might
be
available
or
it
might
not,
as
the
same
with
like
is
yeah
the
nahuatl
project
and
if
you
don't
think
you
might
enjoy
doing
ago,
but
ok
see
you
go
yeah
and.
A
F
A
Yeah,
that's
something
that
needs
looking
into
and
it's
a
bigger
problem
than
just
offsets.
That's
a
general
last
reply,
problem
and
yeah
I
think
we
have
a
particularly
like
a
particularly
extreme
case
of
it.
So
I
transform
staff
is
a
little
bit
complicated
doing
that
and
yeah,
but
yeah
I
mean
it's
on
the
best
object.
Okay,.
J
A
B
J
J
J
A
C
J
A
J
The
other
thing
that
I
was
going
to
have
part
of
this
discussion.
It
is
about
the
logo,
so
I've
made
some
updates
literally
use
any.
A
E
J
Well,
so
I'm
starting
to
think
about
what
kind
of
tutorials
we
want
to
present
and
we'll
definitely
I
want
us
to
do
this
to
sort
of
like
overwhelm
people
a
little
bit.
So
we'll
definitely
do
that,
introduction
to
Sun
pie
and
Python,
but
I
also
thought
it'd
be
worthwhile
to
go
a
little
bit
further
and
show
some
of
the
neat
cool
stuff
and
what
Alex
the
stuff
you
just
showed
would
be
amazing
to
show
people
the
show
like
look.
J
E
A
Should
put
sbd
on
the
board
meeting
agenda
because
Kevin
will
probably
be
there
and
I
know
he's
his
kisses
him
he's
one
of
the
people,
organizing
I,
think
something
yeah
yeah
cuz
cuz
he's
out
folder,
so
yeah
we
should.
You
should
do
that
or
unless
we
get
into
one
of
these.
We
have
one
of
these
designed
specifically
for
the
purpose.
J
J
A
Some
lovely
code
yeah.
Yes
on
that
Alex
we
should
try
and
get
if
you're
going
to
do
that,
we
should
try
and
tidy
up
well.
Doc's
first
tie
the
opcode
a
bit
and
we
can
say
about
my
code
and
then
we'll
actually
like
do
a
release,
get
it
in
the
Sun
pike
Conda
channel
and
make
it
nice
and
easy
to
in
school
and
everything
well.
C
B
A
A
My
point
is
that
well
pretty
this
way,
I
can't
see
Fido
getting
merged
in
in
the
next
month.
Yeah.
Certainly
they
trend
over
the
last
few
months
is
definitely
shown
that
that
seems
to
be
the
case.
I
can't
see
it
happening.
In
fact,
SPD
might
be
a
good
time
to
actually
all
of
us
sit
down
and
hack
on
that
for
a
while
and
but
we
have
coordinates
in
now.
We
have
Python
3
support.
In
now
I've
been
using
python
3
to
dev
for
the
last
few
weeks
since
it
got
merged.
J
J
No
man,
oh
man,
oh
man,
yeah
the
yeah
I-
think
that's
a
good
idea
as
I
I'm.
A
Talking
about
doing
no
point,
7
mid-may
so
that
we've
got
like
maybe
all
of
our
maybe
more
early
May,
beginning
of
the
second
week
in
may
and
a
second
week
in
May,
something
like
that,
so
that
we
got
let
them
we've
got.
That
gives
us
three
or
four
weeks
to
bed
it
in
a
master
and
then
a
couple
of
weeks
to
bed
it
in
as
a
release
before
SVD
yeah,
yeah.
A
C
J
H
J
A
J
One
thing
I
want
to
bring
up
which
I
heard
from
from
Juan
Juan
is
currently
still
poking
around
it.
I
guess
0.5
wherever
we
before
we
put
in
units
and
run.
C
J
Wasn't
really
unclear
what
to
do
so.
He
emailed
me
only
in
kind
of
in
secret
I
think
it
was
maybe
yeah.
He
didn't
want
to
bother
everybody,
but
if
one
didn't
realize
this
and
he's
on
the
board,
it's
actually
voted
on
this
particular
issue.
Then
we
may
want
to
I,
don't
know,
maybe
look
at
our
docs
a
little
bit
more
closely
and
make
sure
that
we
mentioned
this
everywhere
and
I've
certainly
spend
some
time.
Educating
people
have
a
speedy.
That's
a
good
time
to
do
that.
So.
A
Astrup
I
do
have
a
much
more
fancy
changelog
set
up
and
we
do
which
gets
nicely
compiled
into
their
documentation.
Stop
so
might
be
something
that
someone
could
look
up.
Hashtag
not
me.
A
J
A
J
B
A
B
H
H
A
No
I
just
mean
if
we
merge
your
concatenate
function,
we
could
graft
on
multi-day
download
into
the
current
like
Earth
system,
but
yeah,
or
at
least,
if
you
have
your
concatenate
function,
you
can
do
goes
day.
1
plus
goes
day,
two
plus
goes
day
3,
so
we
should
get
nourished.
That
would
be
a
good
start.
That'll.