My Life Forever Changed

I was only about eight. Mom (Betty Harsch) had noticed dryness, chafing and tightness around the webs of her fingers. I remember her looking scared at my father when they discussed it. She saw our family doctor about it, and he, being a small town general practitioner had difficulty recognizing it for what it was. It was 6 months later that she was diagnosed with Scleroderma. She was told by one doctor she had 6 months to one year to live. She lived with it for fifteen years (longer than most) before it finally took her. She had a heart attack. The disease can be systemic and can affect the organs of the body as well, and I guess her heart was the first to go… though she had major problems in her other organs too.

Any disease as crippling as Scleroderma, of course, is horrible and really shouldn’t have to be endured by anyone. But, what some people don’t realize is what it does to families. My father and mother had come from hard working families and had worked hard during their lives, in order to give to their children a better future and to ensure for themselves a decent situation for their retirement years. One would probably consider them upper middle class. Dad was a timber faller specializing in falling trees in hard to get at places or in hazardous situations. Together they owned a retirement home and operated that for 20 years. They had done all the right things in the ways of insuring and protecting themselves with life insurance, home insurance and health insurance. Within 2-3 years, thinking they were well protected, her expensive health care needs would soon hit the cap of what the insurance would cover and that was it, just a letter in the mail. Thank you for 20 years of payments, but you’re on your own now. They would spend the next four years tapping savings, selling the business, selling the assets and finally our home. We ended up in an apartment and on a variety of social welfare programs to get by.

Her suffering and these events forever changed who I am. It has been fifteen years since she died, and I can honestly say there hasn’t been a day gone by I haven’t thought of the disease or the horrible sadness it brought us all.



In this photo you can see her hands, she had no ability to make a fist and could only get thumb and forefinger about an inch from touching, they were little more than useless paddles really. I recall walking in a clothing store with her one year and a woman approaching us stopped some 15-20 feet away, looked at her with her mouth opened and asked quite loudly “My God! What happened to your hands?” My mom replied rather hotly and with even more volume: “The pressure cooker blew up on me!” It would have been hilarious if only we could have forgotten the demon responsible for that interchange. A minute later we left, with her crying softly.

I’ve kept this recollection short and well haven’t divulged too much, you see there are memories for me that are too painful to share here… all I can say is that I wouldn’t wish something like Scleroderma on to my worst enemy. I pray that some day there will be a cure and that no one and no family will ever endure living with this crippling, slowly life draining disease again.

This year, my family and I will participate in the first annual “Stepping Out To Cure Scleroderma.” My four year old daughter and two year old son will walk with me in the 2K fun run. My wife, Lori will run in the 10K timed run. We are asking that our friends, family and contacts find it in their hearts to give as generously as they can by sponsoring us in the event (under my name: http://www.firstgiving.com/timharsch)

My children will never get to meet the spit fire that was their grandmother. But they will learn the spirit of giving by participating in events like these… and hopefully Mom will be with us on that day.

Herbs_Raw15030610 on Flickr - Photo Sharing!

Tonight I scanned in an entire photo album lent to me by my Uncle Herb (my dad's brother and that last living of my Dad's kin). We went to see Herb last summer and I finally got around to scanning the photos. They are still raw, as I'll need to go through them all and crop them out into singles and rotate them and what not, but in any case they are least preserved for now.

This album page has a photo of them next to their motorcycles. Uncle Herb tells me motorcycles are in the blood. I think he's right...
Herbs_Raw15030610 on Flickr - Photo Sharing!

Converting Disconnected Graphs to a Collection of Connected Graphs (revised)

In the original post: Converting Disconnected Graphs to a Collection of Connected Graphs I achieved my goal of converting a disconnected graph into a collection of connected graphs. After some discussion on the JUNG support list I was pointed in the direction of WeakComponentClusterer. Compare this method to the original to see:

* the implementation is now just a few lines of code
* this can be used as a template to provide Collections of other graph types beside just DirectedGraph without having to rewrite the implementation.

Some negatives of this approach over the former:
* it should not perform as well since it first performs some computation to create the vertex set and then performs computation to convert the set into a Collection of graphs. Consequently the set of vertex sets is discarded between the two operations.
* perhaps not a negative but at least notable is the fact that it uses reflection to create a copy of the input graph.

package jung;

import java.util.Collection;
import java.util.Set;

import com.google.common.base.Function;

import edu.uci.ics.jung.algorithms.cluster.WeakComponentClusterer;
import edu.uci.ics.jung.algorithms.filters.FilterUtils;
import edu.uci.ics.jung.graph.DirectedGraph;

public class DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<V, E>
  implements
  Function<DirectedGraph<V, E>, Collection<DirectedGraph<V, E>>> {
 public Collection<DirectedGraph<V, E>> apply(DirectedGraph<V, E> from) {
  WeakComponentClusterer<V, E> wcc = new WeakComponentClusterer<V, E>();
  Set<Set<V>> vertexSets = wcc.transform(from);
  
  return FilterUtils.createAllInducedSubgraphs(vertexSets, from);
 } // end method

} // end class
And here is a simple test class:
package jung;

import java.util.Collection;

import junit.framework.Assert;

import org.apache.log4j.Logger;
import org.junit.Test;

import edu.uci.ics.jung.graph.DirectedGraph;
import edu.uci.ics.jung.graph.DirectedSparseGraph;
import edu.uci.ics.jung.graph.Graph;

public class TestDisconnectedDirectedGraphToConnectedDirectedGraphCollection {
 private final static Logger logger = Logger
   .getLogger(TestDisconnectedDirectedGraphToConnectedDirectedGraphCollection.class);

 @Test
 public void testSubGraphs() {
  DirectedGraph<String, String> dg = DirectedSparseGraph
    .<String, String> getFactory().create();

  // g1: a connected graph
  uniqEdge(dg, "A", "B");
  uniqEdge(dg, "E", "B");
  uniqEdge(dg, "B", "C");
  uniqEdge(dg, "B", "D");

  // g2: another connected graph
  uniqEdge(dg, "F", "G");
  uniqEdge(dg, "F", "H");

  // g3: a pair of nodes connected by an edge
  uniqEdge(dg, "I", "J");

  // g4: a loop
  uniqEdge(dg, "K", "K");

  // g5: a weakly connected graph
  uniqEdge(dg, "L", "M");
  uniqEdge(dg, "N", "M");

  logger.debug(dg);

  DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<String, String> f = new DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<String, String>();
  Collection<DirectedGraph<String, String>> g = f.apply(dg);
  logger.debug(g);

  Assert.assertEquals("union of disjoint graphs is not the correct size",
    5, g.size());
 } // end test

 private void uniqEdge(Graph<String, String> g, String src, String dest) {
  g.addEdge(src + dest, src, dest);
 }

} // end test class
And here is the output of the test case:
2010-02-25 14:47:51,058 [TestDirectedGraphToDisjointUnionOfWeaklyConnectedDirectedGraphs.java 43] DEBUG - [Vertices:K
Edges:KK[K,K] , Vertices:I,J
Edges:IJ[I,J] , Vertices:F,G,H
Edges:FG[F,G] FH[F,H] , Vertices:L,M,N
Edges:NM[N,M] LM[L,M] , Vertices:D,E,A,B,C
Edges:AB[A,B] EB[E,B] BC[B,C] BD[B,D] ]

Converting Disconnected Graphs to a Collection of Connected Graphs

I recently discovered JUNG, and I have a use case to treat all weakly connected subgraphs of a directed graph as separate graphs. So given any directed graph, be it just one connected directed graph or several connected, strongly connected or weakly connected directed graphs this code will return to you the disjoint union of all graphs as a collection of connected graphs with edge direction maintained from the original.

The code uses implements Function from google collections, since it could be useful to apply this code to a collection of input graphs. Also I just prefer google collections over jakarta. It however does use Jakarta collections as well since that is what JUNG uses.

package jung;

import java.util.Collection;
import java.util.List;
import java.util.Set;
import java.util.Stack;

import org.apache.commons.collections15.Factory;

import com.google.common.base.Function;
import com.google.common.collect.Lists;
import com.google.common.collect.Sets;

import edu.uci.ics.jung.graph.DirectedGraph;
import edu.uci.ics.jung.graph.DirectedSparseGraph;

public class DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<V, E>
  implements
  Function<DirectedGraph<V, E>, Collection<DirectedGraph<V, E>>> {
 private Set<V> visited;

 public Collection<DirectedGraph<V, E>> apply(DirectedGraph<V, E> from) {
  List<DirectedGraph<V, E>> subGraphs = Lists.newArrayList();

  Stack<V> verticesRemaining = new Stack<V>();
  verticesRemaining.addAll(from.getVertices());

  Factory<DirectedGraph<V, E>> graphFactory = DirectedSparseGraph
    .<V, E> getFactory();

  do {
   visited = Sets.newHashSet();
   DirectedGraph<V, E> newGraph = graphFactory.create();
   dfsCopy(verticesRemaining.pop(), from, newGraph);
   verticesRemaining.removeAll(visited);
   subGraphs.add(newGraph);
  } while (!verticesRemaining.isEmpty());

  return subGraphs;
 } // end method

 private void dfsCopy(V v, DirectedGraph<V, E> from, DirectedGraph<V, E> to) {
  if (!visited.add(v)) {
   return;
  } // end if

  for (E e : from.getInEdges(v)) {
   V neighbor = from.getSource(e);
   to.addEdge(e, neighbor, v);
   dfsCopy(neighbor, from, to);
  } // end for
  for (E e : from.getOutEdges(v)) {
   V neighbor = from.getDest(e);
   to.addEdge(e, v, neighbor);
   dfsCopy(neighbor, from, to);
  } // end for
 } // end method
} // end class

And here is a simple test class:
package jung;

import java.util.Collection;

import junit.framework.Assert;

import org.apache.log4j.Logger;
import org.junit.Test;

import edu.uci.ics.jung.graph.DirectedGraph;
import edu.uci.ics.jung.graph.DirectedSparseGraph;
import edu.uci.ics.jung.graph.Graph;

public class TestDisconnectedDirectedGraphToConnectedDirectedGraphCollection {
 private final static Logger logger = Logger
   .getLogger(TestDisconnectedDirectedGraphToConnectedDirectedGraphCollection.class);

 @Test
 public void testSubGraphs() {
  DirectedGraph<String, String> dg = DirectedSparseGraph
    .<String, String> getFactory().create();

  // g1: a connected graph
  uniqEdge(dg, "A", "B");
  uniqEdge(dg, "E", "B");
  uniqEdge(dg, "B", "C");
  uniqEdge(dg, "B", "D");

  // g2: another connected graph
  uniqEdge(dg, "F", "G");
  uniqEdge(dg, "F", "H");

  // g3: a pair of nodes connected by an edge
  uniqEdge(dg, "I", "J");

  // g4: a loop
  uniqEdge(dg, "K", "K");

  // g5: a weakly connected graph
  uniqEdge(dg, "L", "M");
  uniqEdge(dg, "N", "M");

  logger.debug(dg);

  DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<String, String> f = new DisconnectedDirectedGraphToWeaklyConnectedDirectedGraphCollection<String, String>();
  Collection<DirectedGraph<String, String>> g = f.apply(dg);
  logger.debug(g);

  Assert.assertEquals("union of disjoint graphs is not the correct size",
    5, g.size());
 } // end test

 private void uniqEdge(Graph<String, String> g, String src, String dest) {
  g.addEdge(src + dest, src, dest);
 }

} // end test class

And here is the output of the test case:
2010-02-25 14:47:51,058 [TestDirectedGraphToDisjointUnionOfWeaklyConnectedDirectedGraphs.java 43] DEBUG - [Vertices:K
Edges:KK[K,K] , Vertices:I,J
Edges:IJ[I,J] , Vertices:F,G,H
Edges:FG[F,G] FH[F,H] , Vertices:L,M,N
Edges:NM[N,M] LM[L,M] , Vertices:D,E,A,B,C
Edges:AB[A,B] EB[E,B] BC[B,C] BD[B,D] ]

NOTE: See this follow on post for a revised version

Pound for Pound Challenge



I just joined the Biggest Loser - Pound for Pound Challenge. I pledged to lose 8 pounds by event end ( I think it's in June ). Anyway, hope this helps me and some family in the community.

Pound for Pound Challenge Home

The Last Carnegie Ride?

Kyle and I went riding at Carnegie (near Livermore,CA) this morning for what may be the last time... I certainly hope not, but they plan to close it, I think, next week, for what might be a permanent closure.

They had most of Corral Hollow Creek closed off (this is the area close to the rode and near the entrance to the park). But, otherwise with the fog keeping the ground a touch wet the traction was great, and the temperature was about 50 degrees. So, the conditions were near perfect. I sure hope they don't close this park for good: So many families have had great times there.

I've always felt that the dirt biking community is a small community, full of hard working folks who can barely speak up for themselves and are often picked on by larger communities like environmentalists. Don't get me wrong I'm more green than many but environmentalists have so many bigger fish to fry. Unfortunately, they have trouble battlign the big boys like gas companies and automobile makers. So, it's just easy to get a few wins against the little guy.

For more info see:
http://www.sharetrails.org/releases/media/?story=671

Vote to keep it open here:
http://carnegieforever.org/

Why engineers and motorcycles don't mix

Wait, I'm an engineer (software engineer :-) and I ride motorcycles... maybe the title of this blog entry should be "sometimes engineers and motorcycles don't mix". That would have applied to me a few times I can think of...