CNF:CNF Data Generation: Difference between revisions

From crtc.cs.odu.edu
Jump to navigation Jump to search
← Older edit
Spyridon97 (talk | contribs)
CNF, external, internal
2,385 edits
Spyridon97 (talk | contribs)
CNF, external, internal
2,385 edits
 
(26 intermediate revisions by the same user not shown)
Line 2: Line 2:


= Repository =
= Repository =
# [https://git-community.cs.odu.edu/crtc/nuclear-physics/cnf-data-generation Gitlab]
# [https://github.com/spyridon97/CNF-Data-Generation Github]
# [https://github.com/spyridon97/CNF-Data-Generation Github]


Line 10: Line 11:
= Code =
= Code =
== Summer 2020 ==
== Summer 2020 ==
# Create a script that builds everything (including VTK library) ✔
# Set up a MySQL server on Wahab for partons ✔
# Create a MySQL server for partons
# Create a container with all the required dependencies ✔
# Create a container on wahab
# Split Mitch's Code into:
# Split Mitch's Code into:
## A program that creates kinematic values per core using:
## A program that splits kinematic values for each core using:
### Uniform Data ✔
### Uniform Data ✔
### Mesh Points ✔
### Mesh Points ✔
### Splitting way
### Splitting way
#### First implementation: split the kinematics dataset like. This way we create '''areas''' of kinematics:
#### First implementation: split the kinematics dataset like. This way we create '''areas''' of kinematics. ✔
##### N0_0... N0_m-1,  
##### N0_0... N0_m-1,  
##### N1_0... N1_m-1,  
##### N1_0... N1_m-1,  
Line 24: Line 24:
##### ...,  
##### ...,  
##### Nn-1_0... Nn-1_m-1.  
##### Nn-1_0... Nn-1_m-1.  
#### Second implementation: split the kinematics dataset by creating chunks of size 1 until each kinematics dataset for each node is filled. This way all nodes will have points of similar kinematics. [<span style="color:green">DONE</span>]
#### Second implementation: split the kinematics dataset by creating chunks of size 1 until each kinematics dataset for each node is filled. This way all nodes will have points of similar kinematics. ✔
##### N0_0, N1_0, N2_0 .. Nn-1_0,  
##### N0_0, N1_0, N2_0 .. Nn-1_0,  
##### N0_1, N1_1, N2_1 .. Nn-1_1,  
##### N0_1, N1_1, N2_1 .. Nn-1_1,  
##### N0_2, N1_2, N2_2 .. Nn-1_2,  
##### N0_2, N1_2, N2_2 .. Nn-1_2,  
##### ...,  
##### ...,  
##### N0_m-1, N1_m-1, N2_m-1 .. Nn-1_m-1## A program that computes the Partons Data per core using the kinematics values and the models:
##### N0_m-1, N1_m-1, N2_m-1 .. Nn-1_m-1
## A program that computes the Partons Data per core using the kinematics values and the models:
### Compton Form Factors ✔
### Compton Form Factors ✔
### Cross-Sections (TODO)
### Cross-Sections (TODO)
## A program that merges the results of each core into one file ✔
## A program that merges the results of each core into one file ✔
## A program that updates the weights of a mesh using partonsData ✔
## A program that updates the weights of a mesh using partonsData ✔
# Create a script to run partons in parallel using shared memory ✔
# Enable shared memory parallelization using CPU affinity, which means pinning a process to a specific core. ✔
# Create a script to run partons in parallel using distributed+shared memory (TODO)
# Create a script that runs partons in parallel using shared memory ✔
# Create a script that runs partons in parallel using distributed + shared memory ✔
# Speedup:
## [[File:partons-speedup.png|500px|File:partons-speedup.png]]
 
== Fall 2020 ==
# Fix Axis ordering. ✔
# Generate Images with spacing. ✔
# Fix max connections problem with the same username on the database. ✔
# Parallelize split and merge steps ✔
# Update VTK to 9.0.1 inside the container. (TODO)
 
== Spring 2021 ==
# Use meshes with double precision weights array ✔
# Add Mitch's code for the computation of error and addition of extra points under the required threshold ✔

Latest revision as of 10:24, 14 April 2021

Repository

  1. Gitlab
  2. Github

Documentation

Summer 2020

  1. Add README.md file ✔

Code

Summer 2020

  1. Set up a MySQL server on Wahab for partons ✔
  2. Create a container with all the required dependencies ✔
  3. Split Mitch's Code into:
    1. A program that splits kinematic values for each core using:
      1. Uniform Data ✔
      2. Mesh Points ✔
      3. Splitting way
        1. First implementation: split the kinematics dataset like. This way we create areas of kinematics. ✔
          1. N0_0... N0_m-1,
          2. N1_0... N1_m-1,
          3. N2_0... N2_m-1,
          4. ...,
          5. Nn-1_0... Nn-1_m-1.
        2. Second implementation: split the kinematics dataset by creating chunks of size 1 until each kinematics dataset for each node is filled. This way all nodes will have points of similar kinematics. ✔
          1. N0_0, N1_0, N2_0 .. Nn-1_0,
          2. N0_1, N1_1, N2_1 .. Nn-1_1,
          3. N0_2, N1_2, N2_2 .. Nn-1_2,
          4. ...,
          5. N0_m-1, N1_m-1, N2_m-1 .. Nn-1_m-1
    2. A program that computes the Partons Data per core using the kinematics values and the models:
      1. Compton Form Factors ✔
      2. Cross-Sections (TODO)
    3. A program that merges the results of each core into one file ✔
    4. A program that updates the weights of a mesh using partonsData ✔
  4. Enable shared memory parallelization using CPU affinity, which means pinning a process to a specific core. ✔
  5. Create a script that runs partons in parallel using shared memory ✔
  6. Create a script that runs partons in parallel using distributed + shared memory ✔
  7. Speedup:
    1. File:partons-speedup.png

Fall 2020

  1. Fix Axis ordering. ✔
  2. Generate Images with spacing. ✔
  3. Fix max connections problem with the same username on the database. ✔
  4. Parallelize split and merge steps ✔
  5. Update VTK to 9.0.1 inside the container. (TODO)

Spring 2021

  1. Use meshes with double precision weights array ✔
  2. Add Mitch's code for the computation of error and addition of extra points under the required threshold ✔
Retrieved from "https://crtc.cs.odu.edu/index.php?title=CNF:CNF_Data_Generation&oldid=8448"