|
Size: 725
Comment:
|
← Revision 3 as of 2015-12-15 12:05:39 ⇥
Size: 1732
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 21: | Line 21: |
| The purpose of this test was to provide a qualitative look into the sensitivity of the SPC iterating process to 3 variables which make up a non-deterministic portion of result uncertainty. These variables are: number of processors specified in [[INIT_LITHOS.TXT]], physical machine structure, different operators. | The purpose of this test was to provide a qualitative look into the sensitivity of the SPC iterating process to 3 variables which make up a non-deterministic portion of result uncertainty. These variables are: number of processors specified in [[INIT_LITHOS.TXT]], physical machine structure, different operators. It is ''very'' important to note that since iterating uses a random seed, it is generally not an exactly repeatable process. This means we can only speak to the qualitative nature of the effect of each the aforementioned variables. Even still, this test gives an idea of the fundamental uncertainty that will be present in SPC results just due to certain uncontrollable environment variables. <<BR>> == Methodology == All members of the SPC team were given an identical starting directory, and were to use identical procedures to build and iterate topography from the given data set. The goal was to tile at 10 and 5 cm GSD, and then iterate the 5 cm tiling 100 times. This data set was comprised of images generated from the same REGION.MAP as all other Test 11 variants, and had viewing conditions as follows: ||<:> Sun ||<:> S/C Az ||<:> S/C Zenith ||<:> Total Number of Images || ||<:> 45, 0, -45 ||<:> every 60 ||<:> 45 ||<:> 24 || <<BR>> == Results == === Processor Comparison === |
Calibration - Test11I
Test Structure
Processors |
Ormacsrv1 |
Ormacsrv2 |
Ormacsrv3 |
DD |
6 |
X |
X |
DL |
X |
8 |
KD, TC |
TC |
X |
JW |
12 |
X |
TC |
X |
X |
Key:
- DL-Diane Lambert
- KD-Kristofer Drozd
- TC-Tanner Campbell
- JW-John Weirich
Purpose
The purpose of this test was to provide a qualitative look into the sensitivity of the SPC iterating process to 3 variables which make up a non-deterministic portion of result uncertainty. These variables are: number of processors specified in INIT_LITHOS.TXT, physical machine structure, different operators. It is very important to note that since iterating uses a random seed, it is generally not an exactly repeatable process. This means we can only speak to the qualitative nature of the effect of each the aforementioned variables. Even still, this test gives an idea of the fundamental uncertainty that will be present in SPC results just due to certain uncontrollable environment variables.
Methodology
All members of the SPC team were given an identical starting directory, and were to use identical procedures to build and iterate topography from the given data set. The goal was to tile at 10 and 5 cm GSD, and then iterate the 5 cm tiling 100 times. This data set was comprised of images generated from the same REGION.MAP as all other Test 11 variants, and had viewing conditions as follows:
Sun |
S/C Az |
S/C Zenith |
Total Number of Images |
45, 0, -45 |
every 60 |
45 |
24 |
Results