The First Cry of Atom

The other day, I gave a presentation at #bq_sushi about Apache Drill. I had totally no experience of using Google BigQuery. But in this year I am very interested in SQL engine as a pipeline in lambda architecture. I have some knowledge and experience of realtime processing and batch layer... [Read More]

Recently I read Apache Drill documents, source code and paper. My major task and interest were focused on realtime processing and Hadoop resource management. So I have no knowledge about adhoc query platforms in spite of the fact that I am a big data developer.So in this chance I decided... [Read More]

Have you ever use Sphinx to create beautiful documentation in your project? Although I have heard about Sphinx and usefulness of it before, this is the first experience for me. In my case, I want to generate an API documentation of my python project. If you create API documentation by... [Read More]

最近英語を使って仕事をすることが多い。もともとソフトウェアのドキュメントは英語で書かれているから読むことは多かった。 でもここ最近はOSSのコミットも結構まじめにやるようになってきたし、毎日会って仕事を一緒にする人の中に日本語が話せない人がちょくちょく増えてきたからだと思う。（今の仕事はとても刺激的だ）昔から英語それ自体を勉強しようと思ったことはあんまりなくて、 必ず何かに使うからというモチベーションでいつも英語を学んできた。もともと本を読むのが好きだったし、世界にある文章のほとんどは英語で書かれていると知ってからはなおさら英語を読めるようになりたいと思ってきた。 [Read More]

Autoencoder can extract various type of features from image sets. As I have done this before with MNIST datasets, we can see this result with our eyes by making images which represent its weight parameter. Today I tried other type of autoencoder which is called MADE(Masked Autoencoder for Distribution Estimation).... [Read More]