Roban Hultman Kramer roban

## AI_policy_panel_IAS.md

      
              2 files
            
          
              0 forks
            
          
              0 comments
            
          
              1 star
            
          
                roban
                / AI_policy_panel_IAS.md
            
            
              Last active
              June 7, 2023 06:31
            
              
                Notes on "Steering AI for the Public Good: A Dialogue for the Future" panel (2023-06-06 at IAS)
              
          
    "Steering AI for the Public Good: A Dialogue for the Future"

2023-06-06 at IAS
Youtube Recording

Panel


## keybase.md

      
              1 file
            
          
              0 forks
            
          
              0 comments
            
          
              0 stars
            
          
                roban
                / keybase.md
            
            
              Created
              March 16, 2016 14:43
            
              
                keybase.md
              
          
    Keybase proof

I hereby claim:

I am roban on github.
I am robanhk (https://keybase.io/robanhk) on keybase.
I have a public key whose fingerprint is 22BD E895 8940 7BBC EEA1  CD3B FE6E A1BB 6A0E 3E55

To claim this, I am signing this object:

  
## index.html
<html>
<head>
<title>Swimlane using d3.js</title>
<script type="text/javascript" src="http://mbostock.github.com/d3/d3.v2.js"></script>
<script type="text/javascript" src="randomData.js"></script>
<style>
.chart {
	shape-rendering: crispEdges;
}

## copy_traces.py
def copy_traces(db1, dbname, dbclass, traceclass, **kwargs):
	"""Copy traces from a db to a new db.
	"""
	db = dbclass(dbname=dbname, dbmode='w', **kwargs)
	db._initialize({}, 1)
	db.trace_names = db1.trace_names
	db.chains = 1
	# Create the Traces.
	for name, trace in db1._traces.iteritems():
		db._traces[name] = traceclass(name=name, value={0:trace[:]}, db=db)

## make_gif.bash
# `convert` is part of Image Magick http://www.imagemagick.org
# The command below scales the images to 500 pixels (in some dimension),
# imposes a delay of 60/100ths of a second between frames, and loops indefinitely.
# The "-layers optimize" option reduces the final file size by optimizing the GIF frames, with a difference threshold
# controlled by the "-fuzz" parameter.
convert -fuzz 2% -layers optimize -scale 500 -delay 60 -loop 0 -verbose IMG_*.JPG animation.gif

## lifetime_mass.py
# paste this into an "ipython --pylab" session
# You'll need EnrichPy: http://roban.github.com/EnrichPy/
import enrichpy.lifetime
mass = logspace(-1,4,300)
tlife = enrichpy.lifetime.main_sequence_life_K97(mass)
plot(mass, tlife)
xscale('log')
yscale('log')
xlabel('mass (solar masses)')
ylabel('stellar lifetime (years)')

## trackpoint_sensitivity.sh
sudo -s "echo -n 250 > /sys/devices/platform/i8042/serio1/serio2/sensitivity"

## makemovie.sh
# first create a list of frames, perhaps by running:
# ls -tr *.JPG > frames.txt

# Now run mencoder to make the move. I adapted the configuration from
# http://blog.hugochinchilla.net/2011/09/time-lapse-videos-mencoder/

# playback at 15 frames per second
mencoder -nosound -ovc lavc -lavcopts vcodec=mpeg4:mbd=2:trell:autoaspect:vqscale=3 -vf scale=1920:1440 -mf type=jpeg:fps=15 mf://@frames.txt -o time_lapse.avi

# playback at 30 frames per second

## plot2Ddist.py
"""Routines related to plotting 2D distributions of parameters.

The core of this module is plot2Ddist, which plots a two-dimensional
distribution with 1D marginal histograms along each axis and optional
features like contours and lines indicating ranges and true values.

"""

import itertools
import math

## plot2Ddist.py
import pylab
import numpy
import pymc
import matplotlib.patches
from mpl_toolkits.axes_grid1 import make_axes_locatable
import scipy.stats

def frac_inside_poly(x,y,polyxy):
    """Calculate the fraction of points x,y inside polygon polyxy.
	<html>
	<head>
	<title>Swimlane using d3.js</title>
	<script type="text/javascript" src="http://mbostock.github.com/d3/d3.v2.js"></script>
	<script type="text/javascript" src="randomData.js"></script>
	<style>
	.chart {
	shape-rendering: crispEdges;
	}
	def copy_traces(db1, dbname, dbclass, traceclass, **kwargs):
	"""Copy traces from a db to a new db.
	"""
	db = dbclass(dbname=dbname, dbmode='w', **kwargs)
	db._initialize({}, 1)
	db.trace_names = db1.trace_names
	db.chains = 1
	# Create the Traces.
	for name, trace in db1._traces.iteritems():
	db._traces[name] = traceclass(name=name, value={0:trace[:]}, db=db)
	# `convert` is part of Image Magick http://www.imagemagick.org
	# The command below scales the images to 500 pixels (in some dimension),
	# imposes a delay of 60/100ths of a second between frames, and loops indefinitely.
	# The "-layers optimize" option reduces the final file size by optimizing the GIF frames, with a difference threshold
	# controlled by the "-fuzz" parameter.
	convert -fuzz 2% -layers optimize -scale 500 -delay 60 -loop 0 -verbose IMG_*.JPG animation.gif
	# paste this into an "ipython --pylab" session
	# You'll need EnrichPy: http://roban.github.com/EnrichPy/
	import enrichpy.lifetime
	mass = logspace(-1,4,300)
	tlife = enrichpy.lifetime.main_sequence_life_K97(mass)
	plot(mass, tlife)
	xscale('log')
	yscale('log')
	xlabel('mass (solar masses)')
	ylabel('stellar lifetime (years)')
	# first create a list of frames, perhaps by running:
	# ls -tr *.JPG > frames.txt

	# Now run mencoder to make the move. I adapted the configuration from
	# http://blog.hugochinchilla.net/2011/09/time-lapse-videos-mencoder/

	# playback at 15 frames per second
	mencoder -nosound -ovc lavc -lavcopts vcodec=mpeg4:mbd=2:trell:autoaspect:vqscale=3 -vf scale=1920:1440 -mf type=jpeg:fps=15 mf://@frames.txt -o time_lapse.avi

	# playback at 30 frames per second
	"""Routines related to plotting 2D distributions of parameters.

	The core of this module is plot2Ddist, which plots a two-dimensional
	distribution with 1D marginal histograms along each axis and optional
	features like contours and lines indicating ranges and true values.

	"""

	import itertools
	import math
	import pylab
	import numpy
	import pymc
	import matplotlib.patches
	from mpl_toolkits.axes_grid1 import make_axes_locatable
	import scipy.stats

	def frac_inside_poly(x,y,polyxy):
	"""Calculate the fraction of points x,y inside polygon polyxy.