Mjesečna Arhiva: Siječanj 2013.

Predictive Analytics and R, Part 2

Data preparing examples: #FROM DATABASE library(‘RPostgreSQL’) drv <- dbDriver(“PostgreSQL”) con <- dbConnect(drv, dbname=”postgres”,user=”postgres”,password=”****”) #Pivoting rs <- dbSendQuery(con,paste(“select t.prdat,”, “coalesce(max(case when t.prioritet=’Emergency’ then t.cnt end),0) as Emergency,”, “coalesce(max(case when t.prioritet=’High’ then t.cnt end),0) as High,”, “coalesce(max(case when t.prioritet=’Medium’ then t.cnt end),0) … Nastavi čitati

Objavljeno u Nekategorizirano | Ostavi komentar

Interoperability

No matter what OS, no matter what browser, no matter what Office suite. The only important thing is the document. And speed.  

Objavljeno u Nekategorizirano | Ostavi komentar

Predictive Analytics and R, Part 1

You need not to believe in your model if your aim is to make a forecast, because e.g. according to BOX AND JENKINS (1976): “All models are wrong but some are useful.” Business decision-making using information such as customer profiles … Nastavi čitati

Objavljeno u Nekategorizirano | Ostavi komentar