View allAll Photos Tagged plot
For any integer x from 1 to 254, I launched a query on yahoo.com on the word yoooooo, with x o's. In other words, I first launched a query on yo, then yoo, yooo, yoooo, yooooo, and so on. Why would I do so? Ask Fabien!
For each such query, I recorded the number of answers found by yahoo, and thus I ended with a number of answers for each value between 1 and 254, which is plotted above (black dots). (I had to stop at 254 because yahoo does not seem to accept queries on longer words).
For the sake of comparison, I did the same thing with various letters in place of o, and other variants. For instance, what I would call the yaaaaa distribution is very similar to the yooooo one. The yxxxxx one, represented on the plot by blue dots, is quite different (significantly below, and maybe sharper slope).
The green line in the plot represents y ~ x^-3 (i.e. the inverse of x to the cube). It fits pretty well the yooooo distribution, showing that it is close to a power-law of exponent 3...
18:00:01 up 10 days, 31 min, 0 users, load average: 0.54, 0.68, 0.72 | temp=42.2'C | Start
18:00:14 up 10 days, 31 min, 0 users, load average: 1.09, 0.79, 0.75 | temp=43.3'C | SID plot Finished
Plot 66: Angela Colleen Clark (5 months) 1966
Brett Kerrigan Clark (stillborn) 10/9/1993
In Loving Memory of
ANGELA COLLEEN
CLARK
31st August 1956 - 20th February 1966.
Beloved daughter of
David and Colleen,
cherished sister of
Paul, Samantha,
Jennyanne and Danielle.
Step softly, a dream lies here.
This concept in Melbourne's northern suburbs - CERES Community Environment Park has been a leader in community building, environmental education and is now embarking on an ambitious and innovative Sustainability Centre - go to www.ceres.org.au
traceroute is a tool that gives the internet path followed by your packets towards a given destination, under the form of a series of IP addresses. Consider a given set of destinations, and let us call the result of a traceroute towards each of these destinations a round of measurement. One may quess that, after a reasonable number of such rounds, no or very few new IP addresses will be discovered by new rounds. One would basically have seen all that can be seen between the computer running traceroute and these destinations.
The plot above shows that this does not happen. We sampled 3000 random IP addresses which answered to ICMP Echo Request and ran a round of measurement every 15 minutes (approximately 100 times a day). We report the number of distinct IP addresses seen since the beginning (in black) and the number of IP addresses which were present at all rounds (in blue) as a function of the number of rounds of measurements we ran.
Surprisingly enough, the number of observed IP addresses still grows after 2100 rounds (3 weeks of continuous measurement). Moreover this growth is not neglectible: in the last 100 rounds, the number of observed IP addresses increases by 2.6 %. Likewise, the number of stable addresses continuously decreases.
One may wonder if this is due to dynamic addresses in our destination set; to check this, we kept only destinations which were present during 90% of the rounds and were always observed immediately after the same IP address. We obtained similar plots.One may then guess that the observed growth is due to routers sending wrong (random?) addresses. However, this would lead to special structures in the graph, which we did not observe. It actually seems that the observed growth is due to the growth of the internet itself.
cover to the latest issue of plots where I have a 6 page story featured inside. A western-horror. All in all I like the printing quality of the publication. sure does look good. Plots has their own flickr page at
Many community detection algorithms are non deterministic and can therefore give different partitions for the same graph. Depending on the context, it can be important to obtain stable results so as to identify very pertinent communities, but it can also be interesting to find some less stable ones.
For non deterministic algorithms, comparing two partitions of a given graph is not so easy. Some parameters can be calculated to estimate the similarity between two partitions: rand index, Jaccard index or the mutual information. However these parameters give only an aggregated value which can be hard to interpret.
In the spirit of the rand index, the plot above shows the similarity between 10,000 computations of communities on the same network, the famous Zachary's karate club. The plot is a distribution of the proportion of pairs of nodes which are in the same group, the point (6213 ; 0.014) for instance means that there is 1.4% of pairs of nodes which are placed in the same community 62% of the time.
A deterministic algorithm would always place nodes either together or not, the curve would therefore exhibit two peaks, one on 0 and one on 10,000. However, the algorithm used (the Louvain method) is not deterministic and therefore some pairs are sometimes grouped and sometimes not. Despite the non-determinism, we can see that most pairs are nearly always grouped or separated, but that around 10% of pairs of nodes are nearly as often together than separated. These nodes are centainely specific and their position have to be investigated.
After a busy session fighting make-believe monsters in the woods, my two young Imps plot their next adventure...
15:00:01 up 13:35, 0 users, load average: 0.51, 0.64, 0.74 | temp=44.4'C | Start
15:00:04 up 13:35, 0 users, load average: 0.51, 0.64, 0.74 | temp=45.5'C | SID plot Finished
Logo concept for Plotter Art. The idea was to highlight CMYK as a means to rpomote their printing business and quality. Thay wanted versatility for different seasons.
Using the data presented in the paper Ten weeks in the life of an eDonkey server we computed the number of queries of each type managed by the server each hour. For each hour, we then computed the ratio between the number of answers to source-search query divided by the number of answers to keyword-driven queries. In other words, we observe the average number of answers to source search queries generated for each answer to a keyword-driven query. We then observed this number for the first hour of all days, second hour of all days, etc, thus obtaining the plot above.
The inset plot was obtained similarily, but by directly considering the number of answers given to keyword-driven queries.
Each blue dot represents an hour of measurement, and the red dots give the average for each hour of day.
In both plots, a clear day/night phenomenom appears, indicating that the large majority of client are in similar timezones (the ones of western Europe, probably).However, the plots have very different shapes, which we may interpret as a consequence of user behavior in reaction to server load (which make it slower).
18:00:02 up 47 min, 0 users, load average: 0.40, 0.64, 0.62 | temp=47.6'C | Start
18:00:16 up 48 min, 0 users, load average: 1.04, 0.77, 0.67 | temp=47.1'C | SID plot Finished
This plot represents the number of papers that were submitted to the French speaking conference Algotel'09. The blue plot represents the number of papers submitted before time t, as a function of time in hours, starting at the first submission. We can clearly see two periods in which there was a high number of submissions (around time 75 and time 175): the blue plot rises sharply at these times. They correspond to the submission deadline for the conference, which was first scheduled on a Friday night, then has been rescheduled for the following Thursday (the blue plot still grows slightly after this second deadline, which corresponds to papers on which modifications were performed after the deadline due to technical reasons).
The red (resp. green) plot represents the number of papers that were submitted before time t and have been accepted for presentation at the conference (resp. rejected). First we can see that there is no overwhelming correlation between the time at which papers were submitted and the fact that they were accepted or rejected: there have been acceptances and rejects for all submission times.
However, we observe that the red plot is above the green one most of the time, which means that, among papers submitted early, a higher fraction was accepted (the final acceptance rate was 48%). We can also observe that, for both deadlines, the red plot rises less sharply than the green one, indicating that papers that were submitted just before the deadline tended to be rejected more often than papers submitted some time before it.
Criação e finzalização do ploter do Ford Ka para o stand do Buriti Shopping
Agencia Jordão Publicidade e Propaganda
It is possible to explore the internet's topology bytracing the paths between some source machines and some destinationmachines.In this way one obtains a subset of this topology. We study here the reliability of the observed properties of this topology,i.e. whether the properties of the subset are properties of the real topology.
In this plot we show the impact of the number of sources and destinations usedon the observed average distance.Each rectangle of coordinates (x,y) in this plot corresponds to a graph obtainedwith y sources and x destinations.The rectangle on the top-left corresponds to 11 sources and 3000 destinations.The color of each rectangle corresponds to the average distance of the correspondinggraph. The grayvariation is linear, from 0 represented by black, to the maximum observed value represented by white.The white line represents the 50% level line, i.e. all points on this line correspondto half the maximum observed value.
In this plot we observe fluctuations for small numbers of sources anddestinations. For instance, with one source the graph is close to a tree, and the average distanceis therefore over-estimated. It changes quickly when only one more source is considered.However, the color becomes uniform once a relatively small number of sourcesand destinations is attained.This shows that the observed average distance does not change much when adding more sourcesand destinations.The observed average distance with our 11 sources and 3000 destinations is therefore probably close to the real value.
Early June 2014, the start of our first full growing season, and the plots are starting to look nice and green.
Two more long-time members of the Once Upon A Time book club for adult readers prepare to share their thoughts on their March selection, The Marriage Plot by Jeffrey Eugenides. Some members enjoyed the coming-of-age tale about a young woman pursing a literature degree in the 1980s while others found her struggles in love tiresome. Many club members said they would recommend it to a friend.
Community detection in complex networks is a hard problem whose classical formulation is the maximisation of the modularity. Since this problem cannot be solved exactly in a reasonable time, heuristics are used to find the best communities.
The Louvain method is an efficient technique to study this problem and consists in a sequence of passes, each being composed of a sequence of iterations. During one iteration of a given pass, all nodes are considered once and are moved from one community to another so as to maximise the gain of modularity. Rather than doing iterations for a given pass while an improvement can be achieved, we study here the loss of quality if we only perform iterations while the improvement is greater than a given epsilon.
The above plot displays the final modularity (in red) and execution time (in blue) as a function of this epsilon. The greater the epsilon, the less iterations are going to be performed at each pass. This can be seen of the blue curve: with a maximal optimization (left part of the curve), the computation time can take as long a 550s, while with a higher value of epsilon, it can be divided by 10. On the other hand, the red curve displays the final modularity obtained and we can see that even with high value of epsilon, the final quality is still good. It is only above a given value, which is between 0.1 and 0.01 depending on the network, that the quality drops suddenly. Therefore, there is not always a compromise to do between time and quality of the decomposition in communities.
In specific cases the modularity is even better with a high epsilon than with a small one. This can be understood as follows: a strong optimisation of the modularity on the first pass can impose too much constraints on the obtained partition and prevent further optimisation to be achieved on subsequent passes. On the other hand, doing a smaller optimization during the first pass can leave much more space for further optimization.
Plot 127: Cyril Desmond Bush
Violet Lena Bush (78) 1980
In Loving Memory
of
CYRIL DESMOND BUSH,
beloved husband of
Violet Lena,
loved father of
Raymond, Collen & Ivan
died 12th June 1974
aged 73 years.
Also his beloved wife & our loved mother
VIOLET LENA
died 6th March 1980
aged 77 years.
BUSH
Since I was a kid, I have always been captivated by the paper shrouded rockets like "The Point". My biggest nemesis, however, was always the Centuri Vulcan. I built lots of them but never got one to behave or even look very good...until I became a BAR and cloned one. It's not the best flyer in the world but it actually works which is more than I can say about any that I built in the 70s.
Recently, I have been in contact with Blades from TRF about his 24mm upscale and he kindly sent me the PDFs for the shrouds. After some conversation, we mentioned other scales and I mentioned that I used to be an engineer and still had a color plotter that handles media up to 5 feet wide and 300 feet long. This also means no splicing of the shrouds!
I tried to print first on a sheet of posterboard but the plotter did not like that. The media was too thick. I visited the local paper distributor and found 130# cardstock that was less expensive, bigger, thinner and stronger than the posterboard. I ordered some sheets and resolved to give it a try.
It took some trial and error but I found that landscape orientation would actually work.