C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
P E T E R S K A N D S : C E R N T H E O R Y U N I T
MonteCarloNet FP7 ITN VINCIA PYTHIA
CERN PILOT PROJECT
LHC Physics Center at CERN
CERN PILOT PROJECT - - PowerPoint PPT Presentation
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P E T E R S K A N D S : C E R N T H E O R Y U N I T CERN PILOT PROJECT
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
P E T E R S K A N D S : C E R N T H E O R Y U N I T
MonteCarloNet FP7 ITN VINCIA PYTHIA
LHC Physics Center at CERN
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Theory Experiment
2
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Theory Experiment
Lots of physics to learn about and play with Lots of hi-tech equipment and fascinating measurements to learn about
2
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Theory Experiment
Lots of physics to learn about and play with Lots of hi-tech equipment and fascinating measurements to learn about
2
→ Feedback to scientists
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
VINCIA PYTHIA …
Simulation Codes
Physics (to learn about) Parameters (to play with) → Simulated Particle Collisions
Experimental Data
Particle Accelerators, Detectors, and Measurements (to learn about) → Published Data Points 3
“Events” “Histograms”
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
4
The true nature of the strong nuclear force is revealed at distances below about 10-15m (= 10-6 nanometers) The energy needed to produce wavelengths that small: kick (accelerate) an electron with 1 billion Volts : 1 Giga-electron-Volt (GeV) The energy of the Large Hadron Collider (LHC) at CERN : 8 TeV The theory - quantum chromodynamics - has a challenging and rich structure, from simple fractals of (quantum) fluctuations within fluctuations, to strongly bound towers of resonances. In high-energy experiments, such as those at CERN, our ability to make forecasts is pushed to the limit, and then some. The task of “event generators” is simple: deliver a complete description, in as much detail as mother nature. The solutions are less simple, and yet the clarity of our vision
“the Terascale” !
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
April 2010
VINCIA
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
from friend of friend. Can a 15-yr old be a one- week intern at CERN?
run-time display for our simulation anyway.
editor to edit input cards. Run-time display to compare output to data.
6
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
from friend of friend. Can a 15-yr old be a one- week intern at CERN?
run-time display for our simulation anyway.
editor to edit input cards. Run-time display to compare output to data.
released as new defaults.
6
Mikkel Jeppsson Plots CERN April 2010
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
Structure and Aims Apologies: quite technical (will return to Content and Learning momentarily)
First Deliverable (M18 - Feb 2014) : Initial Prototype
DoW: “A playable, if crude, game prototype, for early testing and development, accompanied by a report describing its features and per formance characteristics.”
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
8
(To be discussed)
Private job control
Player
Simulation Parameters & Selection of Measurements
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
8
(To be discussed)
Private job control
Player Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results)
Controls Displays Simulation Parameters & Selection of Measurements
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
9
(To be discussed) Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud
Private job control Global job control
Help others when idle:
CYBERLAB
Player
Bookmarks & Annotations Visibility:
Local Global
Controls Displays Simulation Parameters & Selection of Measurements
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
9
(To be discussed) Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud
Private job control Global job control
Help others when idle:
CYBERLAB
Player
Bookmarks & Annotations Visibility:
Local Global
Controls Displays Simulation Parameters & Selection of Measurements
Cloud structure(s)
Collaboration with Imperial
Database(s):
Fully reproducible results (each result also stores versions, card files, etc)
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
9
(To be discussed) Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud
Private job control Global job control
Help others when idle:
CYBERLAB
Player
Bookmarks & Annotations Visibility:
Local Global
Controls Displays Simulation Parameters & Selection of Measurements
Cloud structure(s)
Collaboration with Imperial
Database(s):
Fully reproducible results (each result also stores versions, card files, etc)
UI / Visualization
Based on HTML5 + extra graphics tools? Must be able to annotate every object on it
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
9
(To be discussed) Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud
Private job control Global job control
Help others when idle:
CYBERLAB
Player
Bookmarks & Annotations Visibility:
Local Global
Controls Displays Simulation Parameters & Selection of Measurements
Cloud structure(s)
Collaboration with Imperial
Database(s):
Fully reproducible results (each result also stores versions, card files, etc)
UI / Visualization
Based on HTML5 + extra graphics tools? Must be able to annotate every object on it
Interpolation:
Based on “Professor”
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
10
(To be discussed) Player Player Team Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud Team VM Cloud
Private job control Team job control Global job control
Help others when idle:
TEAM (priority) CYBERLAB
Player
Bookmarks & Annotations Visibility:
Local Team Global
Controls Displays You look at this I look at that Simulation Parameters & Selection of Measurements Overlay Displays
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
11
(To be discussed) Player Player Team Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud Team VM Cloud
Private job control Team job control Global job control
Help others when idle:
TEAM (priority) CYBERLAB
Player
Bookmark & Annotations Visibility:
Local Team Global
Controls Displays You look at this I look at that Simulation Parameters & Selection of Measurements Overlay Displays User mods?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
11
(To be discussed) Player Player Team Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) Cyberlab VM Cloud Team VM Cloud
Private job control Team job control Global job control
Help others when idle:
TEAM (priority) CYBERLAB
Player
Bookmark & Annotations Visibility:
Local Team Global
Controls Displays You look at this I look at that Simulation Parameters & Selection of Measurements Overlay Displays User mods?
User Mods:
Enable users to redefine how the basic numerical data is displayed. Can range from rearranging displays to generating completely new visualizations
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Private computer(s)
12
(To be discussed) Player Player Team Local UI / Visualization Local Interpolation (Changes → Interpolate) Local Database (Simulation Results) Global Database (Simulation Results) LHC@home VM Cloud Cyberlab VM Cloud Team VM Cloud
Private job control Team job control Global job control
LHC@home
job control
Help others when idle:
TEAM (priority) CYBERLAB LHC@home
Player
Bookmarks & Annotations Visibility:
Local Team Global
Controls Displays
Online Reference Plots
You look at this I look at that Simulation Parameters & Selection of Measurements Overlay Displays Overlay Displays
www
User mods?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
User Interface and Storyline Here: focus on user experience and science content Needs to be added (→ platform development?): Hooks for monitoring and evaluating use and learning
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
14
(To be discussed)
interesting and interactive content. Controls Display Task:
… … … … …
i i i i i i i
Rollover tooltips + Click for more More detailed explanations can be clicked into → Explanation → Elaboration → Engagement
Use the controls (left) to make the simulation agree with the data (right)
Engagement Cycle (constructivist)
Note: only for illustration. Will be horribly abused
Start
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
14
(To be discussed)
interesting and interactive content. Controls Display Task:
… … … … …
i i i i i i i
Rollover tooltips + Click for more More detailed explanations can be clicked into → Explanation → Elaboration → Engagement
Controls are used to set the simulation parameters
Use the controls (left) to make the simulation agree with the data (right)
Engagement Cycle (constructivist)
Note: only for illustration. Will be horribly abused
Start
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
15
(Clicked on “Controls”) Controls Display Controls
Are used to set the simulation
… … … … …
i i i i i i i
Rollover tooltips + Click for more Level 1
+ Combine with vote good/bad (incl our explanations) + Forums for further detailed discussion of issues
(can be divided into levels: citizens, students, experts)
user extensions, and discussion
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Nuclear Force
16
(Clicked on “Task” and maximized) Many opportunities for learning + Annotations
(private, team, global)
i i i
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Nuclear Force
16
(Clicked on “Task” and maximized) Many opportunities for learning + Annotations
(private, team, global)
i i i
Learning Monitors User Studies
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Context Example: Theory Simulation
17
Start
The myth tells how Apollon, the God of Wisdom, killed the monster Python, close to the village of Delphi in Greece. To commemorate this victory, Apollon founded the Pythic Oracle, in Delphi, on the slopes of Mount Parnassos…
i i i
+ Annotations
(private, team, global)
Q: is each of these a wiki page, or something else?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Context Example: Theory Simulation
17
Start
The myth tells how Apollon, the God of Wisdom, killed the monster Python, close to the village of Delphi in Greece. To commemorate this victory, Apollon founded the Pythic Oracle, in Delphi, on the slopes of Mount Parnassos…
i i i
+ Annotations
(private, team, global)
Q: is each of these a wiki page, or something else?
?
If so, could this be a dunnonode?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Context Example: Theory Simulation
17
Start
The myth tells how Apollon, the God of Wisdom, killed the monster Python, close to the village of Delphi in Greece. To commemorate this victory, Apollon founded the Pythic Oracle, in Delphi, on the slopes of Mount Parnassos…
i i i
+ Annotations
(private, team, global)
Q: is each of these a wiki page, or something else? Q: is this a knownode?
?
If so, could this be a dunnonode?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Start
18
Context Example: Accelerator
The Large Electron Position (LEP) Collider operated at CERN from 19XX to 20YY. The decision to shut it down was not an easy one, as small hints of the elusive Higgs boson were present in the very last data sets. Were these hints a fluke, or the real thing? Every year of continued running risked delaying the more powerful Large Hadron Collider (LHC), which would replace it. …
i i i
+ Annotations
(private, team, global)
? ?
Q: is each of these a wiki page, or something else?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
Start
19
Context Example: Observable + Annotations
(private, team, global)
Q: is each of these a wiki page, or something else?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
20 “Campaign Mode” Start with a single measurement, and a single parameter to adjust. Gradually work your way up, learning more physics, proceeding to ever more complicated multi-dimensional parameter spaces, with corresponding oodles of experimental measurements. Beat the state of the art and get your name on a new “tune” → the next Jeppsson “Skirmish Mode” Start with a random measurement, and a random simulation. Check if they agree or not. Flag yes/no. Can be useful for validation of new code releases, etc. Can still be used for learning, since the contexts can still be there, and people can click on the measurement, or the simulation setup, … “Science Fiction Mode” Compare to simulated data from a parallel universe … … in which space-time has 5 dimensions, or there are dark- matter particles in the data (or a different Higgs). Or a million other things. Can you tell? Goal is to learn about ideas for new physics and how the real experimenters search for it in the real world (But this is really a separate project in its own right)
(To be discussed)
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
21
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
22
(citizen science) : beat the state of the art → feedback to scientists
Won’t happen every day, and not early. For the patient and the few? Contributing something real to the scientists is main motivator. “Points” and “badges” may make it seem less serious and be counter-productive? (cf. Zooniverse)
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
22
(citizen science) : beat the state of the art → feedback to scientists
Won’t happen every day, and not early. For the patient and the few? Contributing something real to the scientists is main motivator. “Points” and “badges” may make it seem less serious and be counter-productive? (cf. Zooniverse)
(learning) : people can use it just to learn about particle physics. Also useful for outreach, and for physics teaching?
Progress markers may be useful, even desirable. How well am I doing?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
22
(visualization) : scientists also get a nice UI. It then needs to be close enough to the “real deal” that scientists can use it too.
Visual design (plots) must be professional and modifiable, usable in scientific publications. Bonus: can point to same graphics in real science papers
(citizen science) : beat the state of the art → feedback to scientists
Won’t happen every day, and not early. For the patient and the few? Contributing something real to the scientists is main motivator. “Points” and “badges” may make it seem less serious and be counter-productive? (cf. Zooniverse)
(learning) : people can use it just to learn about particle physics. Also useful for outreach, and for physics teaching?
Progress markers may be useful, even desirable. How well am I doing?
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2
Technical Structure and Aims (Too technical to discuss details here, but would like to discuss some of it with experts)
C i t i z e n C y b e r l a b I C T - 1 s t K i c k o f f M e e t i n g , P a r i s , N o v 1 2 - 1 3 2 0 1 2 P e t e r S k a n d s : C E R N P i l o t P r o j e c t
parameter points for which “exact” results exist. Precision on each point limited by amount of generated “events”, which continually increases.
the “professor” tool.
professor’s interpolation(s). Requires request to central database and download of results or local copy of database, periodically updated. His/ her computer then begins further refinement for the particular point chosen, reducing statistical and interpolation errors. (+, optionally, teams can help)
procedure for handling runs with different statistics. Need procedure to determine statistical uncertainty of interpolation + interpolation error.
points to the grid.
parameter by parameter → avoid infinitely many possibilities. 24