This book is intended for education and fun. Python is an amazing, text-based coding language, perfectly suited for children older than the age of 10. The Standard Python library has a module called Turtle which is a popular way to introduce programming to kids. This library enables children to create pictures and shapes by providing them with a virtual canvas. With the Python Turtle library, you can create nice animation projects using images that are taken from the internet, scaled-down stored as a gif-files download to the projects. The book includes 19 basic lessons with examples that introduce to the Python codes through Turtle library which is convenient to the school students of 10+years old. The book has also a lot of projects that show how to make different animations with Turtle graphics: games, applications to math, physics, and science.
SVGo is a Go programming language library for generation of SVG. The talk discusses the design of the library, the concept of sketching in code, and the development of visualizations and tools.
Gentlest Introduction to Tensorflow - Part 3Khor SoonHin
Articles:
* https://medium.com/all-of-us-are-belong-to-machines/gentlest-intro-to-tensorflow-part-3-matrices-multi-feature-linear-regression-30a81ebaaa6c
* https://medium.com/all-of-us-are-belong-to-machines/gentlest-intro-to-tensorflow-4-logistic-regression-2afd0cabc54
Video: https://youtu.be/F8g_6TXKlxw
Code: https://github.com/nethsix/gentle_tensorflow
In this part, we:
* Use Tensorflow for linear regression models with multiple features
* Use Tensorflow for logistic regression models with multiple features. Specifically:
* Predict multi-class/discrete outcome
* Explain why we use cross-entropy as cost function
* Explain why we use softmax
* Tensorflow Cheatsheet #1
* Single feature linear regression
* Multi-feature linear regression
* Multi-feature logistic regression
This book is intended for education and fun. Python is an amazing, text-based coding language, perfectly suited for children older than the age of 10. The Standard Python library has a module called Turtle which is a popular way to introduce programming to kids. This library enables children to create pictures and shapes by providing them with a virtual canvas. With the Python Turtle library, you can create nice animation projects using images that are taken from the internet, scaled-down stored as a gif-files download to the projects. The book includes 19 basic lessons with examples that introduce to the Python codes through Turtle library which is convenient to the school students of 10+years old. The book has also a lot of projects that show how to make different animations with Turtle graphics: games, applications to math, physics, and science.
SVGo is a Go programming language library for generation of SVG. The talk discusses the design of the library, the concept of sketching in code, and the development of visualizations and tools.
Gentlest Introduction to Tensorflow - Part 3Khor SoonHin
Articles:
* https://medium.com/all-of-us-are-belong-to-machines/gentlest-intro-to-tensorflow-part-3-matrices-multi-feature-linear-regression-30a81ebaaa6c
* https://medium.com/all-of-us-are-belong-to-machines/gentlest-intro-to-tensorflow-4-logistic-regression-2afd0cabc54
Video: https://youtu.be/F8g_6TXKlxw
Code: https://github.com/nethsix/gentle_tensorflow
In this part, we:
* Use Tensorflow for linear regression models with multiple features
* Use Tensorflow for logistic regression models with multiple features. Specifically:
* Predict multi-class/discrete outcome
* Explain why we use cross-entropy as cost function
* Explain why we use softmax
* Tensorflow Cheatsheet #1
* Single feature linear regression
* Multi-feature linear regression
* Multi-feature logistic regression
Video: https://youtu.be/dYhrCUFN0eM
Article: https://medium.com/p/the-gentlest-introduction-to-tensorflow-248dc871a224
Code: https://github.com/nethsix/gentle_tensorflow/blob/master/code/linear_regression_one_feature.py
This alternative introduction to Google's official Tensorflow (TF) tutorial strips away the unnecessary concepts that overly complicates getting started. The goal is to use TF to perform Linear Regression (LR) that has only a single-feature. We show how to model the LR using a TF graph, how to define the cost function to measure how well the an LR model fits the dataset, and finally train the LR model to find the best fit model.
Gentlest Introduction to Tensorflow - Part 2Khor SoonHin
Video: https://youtu.be/Trc52FvMLEg
Article: https://medium.com/@khor/gentlest-introduction-to-tensorflow-part-2-ed2a0a7a624f
Code: https://github.com/nethsix/gentle_tensorflow
Continuing from Part 1 where we used Tensorflow to perform linear regression for a model with single feature, here we:
* Use Tensorboard to visualize linear regression variables and the Tensorflow network graph
* Perform stochastic/mini-batch/batch gradient descent
TensorFlow is a wonderful tool for rapidly implementing neural networks. In this presentation, we will learn the basics of TensorFlow and show how neural networks can be built with just a few lines of code. We will highlight some of the confusing bits of TensorFlow as a way of developing the intuition necessary to avoid common pitfalls when developing your own models. Additionally, we will discuss how to roll our own Recurrent Neural Networks. While many tutorials focus on using built in modules, this presentation will focus on writing neural networks from scratch enabling us to build flexible models when Tensorflow’s high level components can’t quite fit our needs.
About Nathan Lintz:
Nathan Lintz is a research scientist at indico Data Solutions where he is responsible for developing machine learning systems in the domains of language detection, text summarization, and emotion recognition. Outside of work, Nathan is currently writting a book on TensorFlow as an extension to his tutorial repository https://github.com/nlintz/TensorFlow-Tutorials
Link to video https://www.youtube.com/watch?v=op1QJbC2g0E&feature=youtu.be
Scientific Computing with Python Webinar March 19: 3D Visualization with MayaviEnthought, Inc.
In this webinar, Didrik Pinte provides an introduction to MayaVi, the 3D interactive visualization library for the open source Enthought Tool Suite. These tools provide scientists and engineers a sophisticated Python development framework for analysis and visualization.
Video: https://youtu.be/dYhrCUFN0eM
Article: https://medium.com/p/the-gentlest-introduction-to-tensorflow-248dc871a224
Code: https://github.com/nethsix/gentle_tensorflow/blob/master/code/linear_regression_one_feature.py
This alternative introduction to Google's official Tensorflow (TF) tutorial strips away the unnecessary concepts that overly complicates getting started. The goal is to use TF to perform Linear Regression (LR) that has only a single-feature. We show how to model the LR using a TF graph, how to define the cost function to measure how well the an LR model fits the dataset, and finally train the LR model to find the best fit model.
Gentlest Introduction to Tensorflow - Part 2Khor SoonHin
Video: https://youtu.be/Trc52FvMLEg
Article: https://medium.com/@khor/gentlest-introduction-to-tensorflow-part-2-ed2a0a7a624f
Code: https://github.com/nethsix/gentle_tensorflow
Continuing from Part 1 where we used Tensorflow to perform linear regression for a model with single feature, here we:
* Use Tensorboard to visualize linear regression variables and the Tensorflow network graph
* Perform stochastic/mini-batch/batch gradient descent
TensorFlow is a wonderful tool for rapidly implementing neural networks. In this presentation, we will learn the basics of TensorFlow and show how neural networks can be built with just a few lines of code. We will highlight some of the confusing bits of TensorFlow as a way of developing the intuition necessary to avoid common pitfalls when developing your own models. Additionally, we will discuss how to roll our own Recurrent Neural Networks. While many tutorials focus on using built in modules, this presentation will focus on writing neural networks from scratch enabling us to build flexible models when Tensorflow’s high level components can’t quite fit our needs.
About Nathan Lintz:
Nathan Lintz is a research scientist at indico Data Solutions where he is responsible for developing machine learning systems in the domains of language detection, text summarization, and emotion recognition. Outside of work, Nathan is currently writting a book on TensorFlow as an extension to his tutorial repository https://github.com/nlintz/TensorFlow-Tutorials
Link to video https://www.youtube.com/watch?v=op1QJbC2g0E&feature=youtu.be
Scientific Computing with Python Webinar March 19: 3D Visualization with MayaviEnthought, Inc.
In this webinar, Didrik Pinte provides an introduction to MayaVi, the 3D interactive visualization library for the open source Enthought Tool Suite. These tools provide scientists and engineers a sophisticated Python development framework for analysis and visualization.
Bestseller Analysis: Visualization Fiction (for PyData Boston 2013)Lynn Cherny
A version of my OpenVisConf talk "Bones of a Bestseller" that gives more detail on topic analysis plus adds python code. Blog post and ipynb code here: http://blogger.ghostweather.com/2013/08/pydata-boston-2013-more-on-fiction.html
Things I Think Are Awesome (Eyeo 2016 Talk)Lynn Cherny
Some text-generation toys I built that use neural nets, genetic algs, poetry. If you want to play with the toys, the link are being cut off on Slideshare at the bottom edge. Go here: https://arnicas.github.io/eyeo2016-talk/
Python is a high level language focused on readability. The Python community developed the concept of "Pythonic Code", requiring not only semantic correctness, but also conformity to universally acknowledged stylistic criteria.
A pre-requisite to write pythonic code is to write idiomatic code. Using the right idioms is a matter of acquired taste and experience, however, some idioms are quite easy to learn.
This presentation focuses on some of these idioms and other stylistic criteria:
* for vs. while
* iterators, itertools
* code conventions (space invaders)
* avoid default values bugs
* first order functions
* internal/external iterators
* substituting the switch statement
* properties, attributes, read only objects
* named tuples
* duck typings
* bits of metaprogramming
* exception management: LBYL vs. EAFP
I am Parton R. I am a Signal Processing Assignment Expert at matlabassignmentexperts.com. I hold a Master's in Matlab, from The University, of Edinburgh, UK. I have been helping students with their assignments for the past 6 years. I solve assignments related to Signal Processing.
Visit matlabassignmentexperts.com or email info@matlabassignmentexperts.com.
You can also call on +1 678 648 4277 for any assistance with Signal Processing assignments.
Volume rendering 3D volume data (medical CT scans) in Unity3D.
Covering the following topics:
- Raymarching
- Maximum Intensity Projection
- Direct Volume Rendering with compositing
- Isosurface rendering
- Transfer functions
- 2D Transfer Functions
- Slice rendering
Source code here: https://github.com/mlavik1/UnityVolumeRendering
DojoX GFX Session Eugene Lazutkin SVG Open 2007Eugene Lazutkin
Eugene Lazutkin's course session on DojoX GFX at SVG Open 2007.
(The keynote is here: http://www.slideshare.net/elazutkin/dojox-gfx-keynote-eugene-lazutkin-svg-open-2007/)
A fast-paced introduction to Deep Learning concepts, such as activation functions, cost functions, back propagation, followed by some TensorFlow features, and then a code sample of training a CNN in tensorflow.js. Basic knowledge of vectors, matrices, and derivatives is helpful in order to derive the maximum benefit from this session.
A fast-paced introduction to Deep Learning concepts, such as activation functions, cost functions, back propagation, followed by some TensorFlow features, and then a code sample of training a CNN in tensorflow.js. Basic knowledge of vectors, matrices, and derivatives is helpful in order to derive the maximum benefit from this session.
Intro to Deep Learning, TensorFlow, and tensorflow.jsOswald Campesato
This fast-paced session introduces Deep Learning concepts, such gradient descent, back propagation, activation functions, and CNNs. We'll look at creating Android apps with TensorFlow Lite (pending its availability). Basic knowledge of vectors, matrices, and Android, as well as elementary calculus (derivatives), are strongly recommended in order to derive the maximum benefit from this session.
A Fast and Dirty Intro to NetworkX (and D3)Lynn Cherny
Using the python lib NetworkX to calculate stats on a Twitter network, and then display the results in several D3.js visualizations. Links to demos and source files. I'm @arnicas and live at www.ghostweather.com.
Design For Online Community: Beyond the HypeLynn Cherny
Reviews academic (anthro, socio, linguistic) definitions of online and offline community, followed by principles for creating them online and measuring their success. (My Ph.D. was an early study of online community. I do data mining/vis now at @arnicas/www.ghostweather.com.)
12. LIBRARIES IN NODEBOX 1 (MAC OSX)
Note:
these
libraries
must
be
put
in
~/Library/Application
Support/Nodebox
to
be
imported.
All
the
libs
live
here.
26. CLOSEST SIMILAR TOOLS
Drawbot
(Preceded
and
inspired
Nodebox,
MacOSX
only)
Shoebot
(MacOSX),
with
Spryte
for
Windows
(some
examples
run
unchanged
in
NB1!)
Pythonista
on
Ipad!
Processing
(cross
platform,
includes
.js
port)
(Processing.py
by
jpheinberg
is
jython-‐based.)
27. PROCESSING LOOKS LIKE JAVA L
Plus,
obviously,
I
want
Python
libs
h3p://openprocessing.org/sketch/8941
31. THE DRAW() LOOP
Nodebox
1
can
be
used
for
simple
static
image
without
animation
–
no
canvas
declaration
or
draw
loop
needed.
(Use
speed(<fps>)
to
turn
on
the
animation.)
Nodebox
OGL
always
runs
an
animation
loop
in
a
draw
function
(you
can
exit
out
with
a
return
after
canvas.frame==1
in
“draw”
if
you
want)
mycanvas = Canvas(width=600, height=480)
mycanvas.fps = 20
mycanvas.run(draw=draw,setup=setup)
33. NOTICE THE CONTEXT AGAIN…
from nodebox.graphics import *
def draw(canvas):
canvas.clear()
nofill() Set
context
values
stroke(0, 0.25)
strokewidth(1) X,
Y,
width,
height
Local
override
of
context
rect( 50, 50, 50, 50) values
rect(110, 50, 50, 50, stroke=Color(0), strokestyle=DASHED)
rect(170, 50, 50, 50)
canvas.run(draw)
34. LEARNING THE “REST”
§ Examples
with
both
NB
1
and
NB
OGL
distribs:
commented
and
by
topic
§ Tutorials
on
the
NB
1
site
§ The
extensive
intro
page
for
NB
OGL
(that
builds
off
NB1’s
api
background)
37. FICTION INVESTIGATION…
Shane
Bergsma’s
db
of
noun
gender
(based
on
Google
news
crawling):
[see
refs]
“word
male
female
neutral
plural”,
e.g.:
publication
93
20
3152
110
1. Load
Shane’s
db
into
redis
2. Convert
books
to
txt
(blank
line
bw
paragraphs)
3.
Extract
nouns
with
pattern.py
4. Code
each
with
tuple
(m,
f,
n)
&
%’s
5. Write
out
as
csv
for
use
in
Nodebox
scripts
39. FOOTNOTE: HSV IN THE BLUE-RED
RANGE / WITH DARKNESS
Code
borrowed
from
an
example
on
StackOverflow
–
tuned
to
get
only
hue
from
blue
to
red
from
complete
HSV
range
40. GET CARTESIAN X, Y COORD FROM A
TUPLE
def to_cart(triple):
(m, f, n) = triple
x = ( f + n / 2.0)
y = math.sqrt(3) * n / 2.0
return x, y
Code
in
my
common.py
file
41. INTERPOLATION
You
often
need
to
map
from
a
data
range
to
another
range
(of
pixels,
or
color
points…).
Mapping
my
X
and
Y
to
colors:
from scipy.interpolate import interp1d
hue_scale = interp1d([0,1],[.67,1])
For
pythonic
hsv
color
and
then
nodebox
rgb:
hsv = (hue_scale(x)[0], 1, 1-y[0])
rgb = Color(colorsys.hsv_to_rgb(*hsv))
I
am
flipping
the
V!
43. EVENTS : LAYERS, MOUSE, KEYS
Layers
in
NB
OGL
are
one
good
way
you
might
handle
“mouseover”
functionality
Layers
have
their
own
draw()
functionality,
and
the
canvas
knows
that
layer
is
in
focus
(under
the
mouse,
via
canvas.focus)
Mouse
events
are
also
handled
nicely
by
canvas.mouse
–
mouse.x,
mouse.y,
etc.
are
available
See
my
example
triangle_layers.py
48. A FAILED EXPERIMENT CAN STILL BE
FUN… ADDING ANIMATION.
Angels
&
Demons
(Brown)
Twilight
(Meyer)
“jade”
Pride
&
Prejudice
(Austen)
The
Secret
Agent
(Conrad)
triangle_bar_uniq.py
57. HOOKING UP OTHER PYTHON LIBS
1. Load
a
book
into
redis
by
line
#
2. Plot
dialog
vs.
exposition
in
a
simple
colored
bar
3. Use
the
redis
db
to
see
what’s
what
in
the
book
on
rollover!
Simple,
and
very
fast!
58.
59. DIALOG TO EXPOSITION…
Twilight
Angels
&
Demons
Secret
Agent
Pride
&
Prejudice
Moby
Dick
(Meyer)
(Brown)
Para
groups:
7
Quote_bar_nodb.py
61. WHY OR WHY NOT NODEBOX?
Advantages
§ Data
as
“art”
–
not
supported
by
Matplotlib
(or
future
ggplot2
ports
to
python)
§ Data
“sketching”
–
speedy
unstructured
pics
§ Animation
is
basic
§ Events
come
along
too
§ You
get
to
write
in
Python
(unlike
w/
Processing)
§ So
you
can
use
other
Python
libs
62. BUT…
§ No
3d
(unlike
matplotlib)
§ PDF
or
SVG
Export
are
required
for
good
print/reuse
(available
in
NB
1,
not
in
NB
OGL
yet)
§ No
web
embedding
/
js
version
(unlike
processing.js)
§ Can’t
use
with
IPython
notebook
(yet)
§ Challenge
of
other
python
libs
with
NB
1
-‐
sad
PYTHONPATH
problem
in
Nodebox
1
(see
appendix
for
tips)
§ Authors
in
Leuven
more
focused
on
NB
3/Pattern.py
than
on
NB1
/
OGL
versions.
Can
we
invigorate
Nobebox
OpenGL?
§ A
general
lack
of
code
examples
to
draw
from…
hopefully
mine
will
help!
63. THAT’S IT - A BIG THANKS!
@deepfoo
for
the
reminder
of
Nodebox1,
Tom
De
Smedt
and
Frederik
De
Bleser
for
email
help,
@minrk
for
help,
@jsundram
for
code
cleanup
advice
(not
all
of
which
I
took),
@pwang
and
#PyData
for
having
me
64. Find
me
@arnicas,
www.ghostweather.com
blog
GET THE CODE FILES HERE!
PDF OF THESE SLIDES HERE.
Apologies
for
the
import
*
and
the
globals…
I
was
following
some
suggesQons
in
the
demos
I
looked
at
which
may
not
have
been
ideal.
65. REFERENCES
§ JanWillem
Tulp
Ghost
Counties
images:
http://www.flickr.com/photos/janwillemtulp/sets/72157626612248205/
§ Code
for
ternary
plots
in
python
and
excel:
http://sourceforge.net/projects/wxternary/
and
Will
Vaughn’s
at
http://wvaughan.org/ternaryplots.html
§ Nodebox
flickr
gallery
§ Running
Nodebox
1
from
command
line:
http://nodebox.net/code/index.php/Console
§ Pattern.py
by
Tom
de
Smedt
(a
Nodebox
original
author)
§ Nodebox
authors
Tom
De
Smedt
and
Frederik
De
Bleser
in
Belgium
§ Shane
Bergsma
and
Dekang
Lin,
“Bootstrapping
Path-‐Based
Pronoun
Resolution,”
In
Proceedings
of
the
Conference
on
Computational
Lingustics
/
Association
for
Computational
Linguistics
(COLING/
ACL-‐06),
Sydney,
Australia,
July
17-‐21,
2006.
(page
w/
db)
66. APPENDIX: NODEBOX 1’S IMPORT PATH
Custom
path,
includes
its
own
python
(64
bit)…
so….
– You
can
install
your
packages
into
NodeBox’s
path,
ie.,
~/
Library/Application
Support/NodeBox/
—
meaning
that
you
can
use
them
from
NodeBox,
but
not
from
other
scripts…
– You
can
import
sys
in
your
NodeBox
code
and
manually
modify
the
sys.path
value
to
add
your
existing
packages…
– You
can
install
packages
into
your
system
site-‐packages
directory,
and
sym-‐link
them
from
NodeBox’s
directory…
– You
can
make
NodeBox
use
your
system
packages
instead
of
it’s
own
by
sym-‐linking
~/Library/Application
Support/
NodeBox
to
your
site-‐packages
directory
of
choice
(ex.,
/
Library/Python/2.5/site-‐packages)
– Some
flavor
of
above
plus
VIRTUALENV
Tips
from
h3p://www.eriksmar3.com/blog/archives/747
Thread
here
too:
h3p://nodebox.net/code/index.php/shared_2008-‐09-‐04-‐01-‐42-‐29
67. APPENDIX: NODEBOX1 AT COMMAND
LINE…
§ Instructions
and
samples
here:
http://nodebox.net/code/index.php/Console
§ Best
to
use
a
virtualenv
again
68. THE NODEBOX “FAMILY”
Nodebox
1
Platform
&
“style”
Mac
OSX
only
(kind
of
Status
No
longer
in
dev,
spotty
URLs
Mac
OSX
Lion
file:
Lion)
–
write
python
code
archiving
online
https://secure.nodebox.net/
(the
original)
downloads/
in
a
simple
IDE
NodeBox-‐1.9.7rc1.zip
Home:
http://nodebox.net/code/
index.php/Home
Github
copy
of
svn
source:
https://github.com/nodebox/
nodebox-‐pyobjc
Nodebox
2
Mac
OSX
–
python
visual
GONE!
Apparently
was
Home:
(the
disappeared)
programming
“blocks”
slow
and
confusing?
http://Beta.nodebox.net
Nodebox
3
Mac
and
Windows
–
no
Not
so
interesting
to
me:
I
Home:
IDE,
no
python
exposed,
want
to
write
python
http://nodebox.net/node/
(the
current
beta)
all
visual
programming?
code.
Nodebox
Mac
and
Windows
–
write
Not
up
to
date
with
Home:
plain
python
code
Nodebox
1
yet
(e.g.,
lack
http://www.cityinabottle.org/
OpenGL
(the
nodebox/
incomplete)
of
libraries,
lack
of
Github
code:
functionality;
not
so
well
https://github.com/nodebox/
documented);
can’t
run
in
nodebox-‐opengl
IPython
notebook
due
to
probable
multithreading
issue(s)