Differences between revisions 2 and 5 (spanning 3 versions)

Block 5 - Iterate V2

Description

This block iterates a set of landmarks specified by the user, using the lithosP.

Procedure

Set-up

Ensure make_scriptP.seed is linked to the script you want to run. For starters, you can use one of the default ones. If needed, you can craft something specific and place it in lsupport.
Run make_lmrklist (Any landmark that is not in LMRKLISTX.TXT will not be added to images!!!!)

  relink script/Piterate1ShSt.seed make_scriptP.seed
  MAKE_LMRKLISTX

Make List of Landmarks to Process

The list of landmarks you want to process is taken from make_script.in. Typical options:

An entire set of landmarks (copy LMRKLIST.TXT)
Just the landmarks used to make a BIGMAP (copy over USED_MAPS.TXT)
Just the newest landmarks created (copy overLMRKLIST1.TXT )

  cp LMRKLIST.TXT LIST.TXT
  duplicates 
  MAKE_LMRKLISTX 
  make_scriptP | tee run.sh

Run the Batch Iteration

  sh run.sh

Monitor Progress

Typically, you use find_nofitP to see the errors and progress. Additionally, you can use monitor to control the start, stop, looping and notification of the iterations.

  ls -l *.INN *.OOT
  find_nofitP

Evaluate Final Results

Use find_nofit to get redo.txt, which is good to use for notes and tracking of the corrections tracking. iterateEval.sh provides more detail on number of images have a problem (both correlation and overlaps). Fix the errors to a point you have a stable solution.

  find_nofitP 
  cat redo.txt >> notes
  iterateEval.sh

Finish the Block

Once errors have been removed, then you can finish the block. If you want to rerun things with different parameters, then finish this block and start over.

  echo 8 | blockFinish Ap6

(Compiled by EEP)

-  ⇤ ← Revision 2 as of 2018-02-01 15:38:02 → 
  Size: 1697
  Editor: JohnWeirich
  Comment:
+   ← Revision 5 as of 2018-11-15 10:55:37 → ⇥
  Size: 1970
  Editor: EricPalmer
  Comment:
-Deletions are marked like this.
+Additions are marked like this.
 Line 1:
-## page was renamed from Block 5 - Iterate V0
-Line 9:
+Line 8:
-== Assumptions ==
 * You have invoked a Bourne-Again Shell (bash)

 * You have completed the processing steps in Block 1 - Setup Directory
-Line 15:
+Line 10:
-Make sure make_scriptP.seed is correct.
-Line 17:
+Line 11:
+=== Set-up ===
-Line 18:
+Line 13:
-=== 3. Set-up batch iteration files ===
+ * Ensure make_scriptP.seed is linked to the script you want to run.  For starters, you can use one of the default ones.  If needed, you can craft something specific and place it in lsupport.
-Line 20:
+Line 15:
-==== Scenario - Full list of landmarks ====
+ * Run [[make_lmrklist]] (Any landmark that is not in [[LMRKLISTX.TXT]] will not be added to images!!!!)
-Line 23:
+Line 18:
-# Set-up batch iteration files
## SCENARIO 1: Iterate all landmarks lised in LMRKLIST.TXT
cp LMRKLIST.TXT LIST.TXT
duplicates 
## END OF SCENARIO 1
}}}

==== Scenario - Recently Tiled ====
{{{
## SCENARIO 2: Iterate landmarks that were just created by tiling 
cp LMRKLIST1.TXT LIST.TXT
duplicates 
## END OF SCENARIO 2
+  relink script/Piterate1ShSt.seed make_scriptP.seed
  MAKE_LMRKLISTX
-Line 39:
+Line 23:
-==== Scenario - List of landmarks from a region (BIGMAP) ====
{{{
## SCENARIO 3: Iterate landmarks which overlap a region-of-interest defined via a bigmap .in file
bigmap < lsupport/<bigmap>.in
cp USED_MAPS.TXT LIST.TXT 
duplicates
## END OF SCENARIO 3
}}}

==== Finish Setup ====
+=== Make List of Landmarks to Process ===
The list of landmarks  you want to process is taken from [[make_script.in]].  Typical options:
 1. An entire set of landmarks (copy [[LMRKLIST.TXT]]) 
 * Just the landmarks used to make a BIGMAP (copy over [[USED_MAPS.TXT]])
 * Just the newest landmarks created (copy over[[ LMRKLIST1.TXT]] )
-Line 51:
+Line 30:
-MAKE_LMRKLISTX 
make_scriptP | tee run.sh
}}}

=== 4. Invoke batch iteration ===
{{{
# Invoke batch iteration
sh run.sh
}}}

=== 5. Monitor progress ===
{{{
# Monitor progress
ls -l *.INN *.OOT
find_nofitP
}}}

=== 6. Check results ===

{{{

# Check results
find_nofitP | tee output
mkdir ../jrwlog/step1-5
mv run_script*.b *INN *OOT output run.sh eval* ../jrwlog/step1-5
+  cp LMRKLIST.TXT LIST.TXT
  duplicates 
  MAKE_LMRKLISTX 
  make_scriptP | tee run.sh
-Line 79:
+Line 37:
-=== 8. Clean directory ===
+=== Run the Batch Iteration ===
-Line 81:
+Line 40:
-# Clean directory
sh rem_script.b
rm TESTFILES/*
rm TESTFILES1/*
+  sh run.sh
}}}

=== Monitor Progress ===
Typically, you use [[find_nofitP]] to see the errors and progress.  Additionally, you can use [[monitor]] to control the start, stop, looping and notification of the iterations.
{{{
  ls -l *.INN *.OOT
  find_nofitP
}}}

=== Evaluate Final Results ===
Use [[find_nofit]] to get [[redo.txt]], which is good to use for notes and tracking of the corrections tracking.  iterateEval.sh provides more detail on number of images have a problem (both correlation and overlaps).   Fix the errors to a point you have a stable solution.
{{{
  find_nofitP 
  cat redo.txt >> notes
  iterateEval.sh
}}}



=== Finish the Block ===
Once errors have been removed, then you can finish the block.  If you want to rerun things with different parameters, then finish this block and start over.
{{{
  echo 8 | blockFinish Ap6
-Line 89:
+Line 68:
-''(Compiled by JRW)''

CategoryHowTos
+''(Compiled by EEP)''