A leaflet.js based interactive map of the coffee shops in NYC that I have visted. The pop-up has the name and my short review/ancedote about the coffee shop. Geographical cordinates are requested from http://mondeca.com/index.php/en/any-place-en
Janu Verma Jverma
de Bruijn graph constructed from a segment of DNA sequence. We used 5-mer dbg which means that the nodes are all the 4-mers (subsequences of length 4) and there is an edge between two nodes is a 5-mer which joins them. These very very useful in genome assembly. This is made using the NetworkxD3 package.
Parallel coordinates visualization of Cosmological parameters which characterizes the chosen cosmological model. There are 38 comsological models in this visual. These models were chosen to built a cosmological emulator, and we used this in our upcoming paper on computation of non-linear matter power spectra.
Visualizing the results of a multpiple DNA sequence alignment. 15 sequences were aligned using ClustalW2 and their results which are of following form -
ATGGAATGATCCATGGT-GCTGACAGATG-CTTCCATGGTGTTAACAGAGTTTCTGGTCT
AAGGGCAAATACATCAT-GAGATGAGTTGGCTTCCATAACATTATAAAACATTTTGATCA
----------AAAACAT-GCAAGAAATGGCTGAACAGGGCACTTTCGGCGG--CAGG---
----------AAATCAT-GCATACAAG---TTCCTAGGGGCCTCAAAACAT--CAAA---
GCCGGGTATTACACCATTGCTGGATGCAATCAAACGAGTGTCTCACTATGCTAAATTTCT
ATAGAGCTGTATAAGGTCCTGAATTGTTCATTCAAGAACTGAAGTCAATGATAACAATCT
Attempt at understanding Hive Plot and creating a D3 visual. Mouseover at each node shows the name and type of the node.
"Hive plots define a linear layout for nodes, grouping nodes by type and arranging them along radial axes based on some property of data. The explicit position encoding has the potential to better reveal the network structure while communicating additional information." - Mike Bostock's blog on Hive Plots.
K-means clustering of the tweets extracted using search API.
Scatterplot of sentiment scores of tweets extracted using search API.
The sentiment is calculated using AFINN scores. For more information see Twitter Sentiment Analysis
It can be seen that a lot of the tweets are neutral (zero sentiment score).
Interactive line chart showing the time series of the total sentiment scores of the tweets realted to a topic extracted using search API. Sentiment on a day is the average of the sentiment scores of all the tweets realted to the given topic extracted that day.
I wrote a blog post for packtpub describing the K-means clustering procedure. The codes I provided were written in R, I also pointed to another blog post of mine where I use python to explain this algotithm. I decided to write a D3 visualization of the clusters.
Detecting communities in the given graph usng Girvan-Newman Algorithm I wrote a python script to extract the communities.