Minha mãe já não está mais tomando sangue, mas precisamos repor o banco de sangue, pediram 15 doadores, mas até o momento conseguimos apenas 7.
Sei que não é das coisas mais motivadoras do mundo ir doar, mas a sensação de bem estar e o senso de haver ajudado é enorme :)
Endereço:
Rua Maestro Cardim, 1041- Bela Vista (próximo a Beneficência Portuguesa)
Em nome de Maria Aparecida Procópio Arjones
De seg. a sex. das 8 às 17 / Sábado das 7 às 14
Muito obrigado!
Beijos
Decided I needed a pre-commit hook for Tumblr while working on the main code base. The pre-commit hook below will automatically lint check all of your php files and bail out if you’re trying to commit broken code.
#!/usr/bin/env bash git diff --cached --name-status | awk '{print $2}' | grep -e...
Last week saw the announcement of Amazon Web Services NoSQL database as a service, Dynamo. Dynamo has a number of very interesting features, not least that it runs as a managed service with certain guarantees with regards to scalability, resilience and performance. From the pricing structure, I…
Download Titan here.
Installing R
$ homebrew install r
$ r
> install.packages(“igraph”, lib=”~/.R/library”)
> library(igraph)
> g<-read.graph(file=’data/graph-of-the-gods.xml’, format=’graphml’)
> plot(g, layout=layout.fruchterman.reingold, vertex.size=4, vertex.label.dist=0.5, vertex.color=”red”, edge.arrow.size=0.5)
Tip: You can automatic include installed libraries in your R session:
$ echo ‘.libPaths( c(“~/.R/library”, .libPaths()) )’ » ~/.Rprofile
Sometimes I wanna to “cache” a result from a very long pipe in bash and once I try to read it again the response from ECHO where all lines concatenated to a single line.
Considering a file with content like:
— simple.log —
1
2
3
4
5
6
My giant pipe command here: :)
$ numbers=`cat simple.log`
If I try something like below it doesn’t work
$ echo $numbers | grep 2
1 2 3 4 5 6
The trick is use ” around the variable and you get the expected result
$ echo “$numbers” | grep 2
2
Excelent post about human-machine cooperation
Analytics: It’s not just about you!
Data needs a schema
Thrift , Avro, etc: Type-safe fails fast with serialization + sane evolution path
Data discovery service
If schema is headmap, metadata is mine lightning.
Reduce cross talk
HCatalog
Global metadata service Maintains nested record abstraction Simplifies format migrations (metadata available, ie: compress algoritm, etc)
DEPENDENCY MAP for free!
Keep costs visible
Otherwise your users will keep asking more, and you’ll need to buy more servers Helps to focus on resources that maybe need to be optimized or your users really need and are important!
Tooling: MR analytics: Cloudera Manager
Build tools not framework: - ppl can use freeze versions,configure and just use it.
Take away
(Source: thestrangeloop.com)
READ: http://jeffjonas.typepad.com/jeff_jonas/2011/04/data-beats-math.html
data + math => better decisions
Analogy with building, you can paint the wall (code feature), but it’s almost impossible to change the core (architecture)
System self-organizing
“When DATABASE is not happy, anyone is happy”
Principles | Incremental scalability and Fault-tolerance
Building Stateful Clustered Services on the JVM
https://github.com/boundary/ordasity
Zookeeper Paper: http://static.usenix.org/events/usenix10/tech/full_papers/Hunt.pdf
Balancing your cluster as HW capacity, not only using EvenlyWorkDistribution
@cscotta | s@boundary.com
(Source: thestrangeloop.com)
“Well behaved system, that has a chance to scale, should NEVER be blocking”
Actor Became
Akka Cluster (experimental)
(Source: thestrangeloop.com)
Memory/interpretation is correlated to TIME, a sequence of images, words, etc
Sparse Distributed Representation (SDRs)
Automated moduel creating Continuos learning Temporal and spatial models
https://www.numenta.com/faq.html
(Source: thestrangeloop.com)
Lack human fault-tolerance
worst thing ever: data loss, data corruption otherwise you can restore system to previous version
Immutability
Conflation of data
Disassociated
Schemas done wrong
function(data unit) => is it valid?
Apache Thrift as schema tool
Most NoSQL keep mutability and it’s not the right direction
What data system do?
query function(all data)
All data —-> Precomputed view —-> Query
f() f()
how to compute views?
lock all data -> Map Reduce (arbitrary functions on arbitrary data)
Should write to a DB that: - is batch write - fast random read - elephantDB or Voldmort
All data must be normalized!
Batch view can be denormalized!
Eventually consistency in a couple of hours, for example.
how to compute query?
Precompute realtime view
Performance and Accuracy problem can coexists
(Source: thestrangeloop.com)
Distance functions
http://books.google.com/books/about/EncyclopediaofDistances.html?id=LXEezzccwcoC
Similar code finder
https://github.com/amuller/checkero
SimMachines: http://simmachines.com/Products/index.html
(Source: thestrangeloop.com)
Authentication
Frameworks
Bishop for restful API in clojure: http://twitch.nervestaple.com/2012/05/07/bishop/
Restlet: http://www.restlet.org/
Discussing group:
Book Refs:
http://www.amazon.com/The-REST-Design-Handbook-ebook/dp/B00890OBFI
(Source: thestrangeloop.com)
Development, Technology, Marketing are among my interests.
