Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31
At 17:00:48, I bought the following two book from bookshop Broekhuis for € 5.00 each:
I came up with a simple trick in the
program to generate more random Street Tile
Patterns. I was not happy with first result,
because it had a strong tendency to place large tiles under eachother with a
one shift to the left. This resulted in many similar looking patches. The trick
was to generate the pattern for the following lines from right-to-left. This
results in a quite even distribution of the different types of tiles. The
reason why the trick works, is that the top-right letter of the tile defines
the tile, and generating from right-to-left, results in a even distribution of
the possibilities when ever there is a choice between the different types of
But on second look, I discovered that the trick does not work perfect, because
the above pattern has several H shapes (rotated by 90°). This does not
completely come as a surprise, but I right now, I have no clue how to fix it.
Random Street Tile Pattern
I found a way to prevent the errors. I realized that there were squares of
four values that did not match the allowed tiles. I added a check to
the program enforcing this. Still there
are some problems at the borders, which I hide by showing only part of them.
More research will be needed to address these issue.
Persistent 'balanced' tree
In the past weeks, I worked on the implementation of a persistent tree data structure. I first wanted to base it on
tree, but then decided on a kind of 'balanced' trees using an extra
height indicator. The idea is that this would also allow less optimal height
differences in favour of sharing larger parts between versions, by only
performing the balancing in the parts that made version specific. This
afternoon, I finished a first version, which works correctly and keeps the
tree reasonable balanced under deletion, but does not implement the idea of
sharing larger parts between versions. I still would like to perform some
performance analyses to see if it can be optimized. All the code is in
a single file, which also
includes all the testing procedures and some debugging code.
You are what you do
This morning, I finished reading the Dutch book
Je bent wat je doet (You are what you do) by Roos Vonk, which I started reading on December 20,
2017, the day I bought the book. Roos Vonk is a Dutch professor of social
psychology. What I like about the book, compared to other self-help books, is
that it contains many references to scientific experiments to support the
advice that it gives. It is also a very practical book. Although I have
finished reading the book, I am going to study again, trying to make a summary
of the main points in order to apply them to my own life.
The Four colour theorem states that faces of a
planar graph can be coloured with four colours. Given four colours, there are
24 ways to permute them. Because of this, there must be another method to
represent a colouring with less freedom. For each planar graph with degree
three it is possible to define an equivalence relationship between a face
colouring and an edge colouring with three colours. Given three colours, there
are six ways to permute them. Which gives reason to believe there is a more
compact way to represent a colouring. And there is indeed, because with each
edge colouring it is the case that at each vertex the three edges are assigned
a different colour. By assigning a number to the edge colours, the colours
will appear either clockwise or anti-clockwise around each vertex. This can
be represented by assigning one of two 'colours' to each vertex. One could
exchange these, meaning that there must still be a more compact method for
representing a face colouring of a planar graph with degree three. This can
done by marking the edges for which the edge colours around the vertices of
the edge run in a different direction. It appears that for a marked edge,
the colours of the faces on the 'opposite side' of the two vertices are always
the same, while for an unmarked edge they are always different. The edge
marking also has the property that each face has an even number of edges with
a marking. But that is about everything that can be stated in general.
Depending on the number of edges of a face, there are certain patterns of
edge markings that are allowed, but there are no simple rules to describe
which patterns can occur.
This evening, I finished reading the book
Philosophy of Andy Warhol by Andy Warhol, which I started reading on
December 9. 2017 after I bought it on November 26.
I found this a rather boring book. It is rather similar to A, A novel in the sense that it present every day talk as something
special. It has long transcripts of telephone calls with B, who according
to Pat Hackett is Brigid
Berlin. It is definitely not about philosophy, at most about Warhol's
attitude to life.
This morning, I went to get my new bike, a Vilo de ville B 200 in the colour apple green. Instead of a chain drive
it has a belt drive, which is more durable, maintenance free, very silent, and
has less friction.
Around seven in the morning, I saw some snow on the
cars, but not on the ground. Around eight, I saw that it was snowing. When
went outside around half past eight, it was snowing. I measured 1.5 cm of snow
on the table outside. Around ten, the snow already started to melt. It snowed
on and off during the day. In the evening there were still some patches of
snow, mostly on (high) grass.
This months interesting links
| December 2017