The Top 10 New Features in KNIME 2

25
The Top 10 New Features in KNIME 2.8 Rosaria Silipo KNIME.com AG, San Francisco

Transcript of The Top 10 New Features in KNIME 2

The Top 10 New Features

in KNIME 2.8

Rosaria Silipo

KNIME.com AG, San Francisco

KNIME 2.8

• KNIME 2.8 was out end of July 2013

• Many New Features

• Documentation available at:

http://tech.knime.org/whats-new-in-knime-28

• My Top 10 New features in KNIME 2.8

Current Date/Time

Current

Date/Time in

Time Generator

Node

New WebPortal

• New Look and Feel

• URLs to run workflows with

parameters

• Possibility of nested workflow calls

New WebPortal

Step 1Select workflow

to run

Step 3, ...More workflow

parameters

Last Step See results

Step 2Run settings and

first workflow parameters

URL Addresses for WebPortal

YouTube video:

“What is new in KNIME WebPortal 3.7” https://www.youtube.com/watch?v=BHrBDo1EIQg

Workflow

Path

Flow Variable

Values

Email

address

Auto-run

parameter

Improved Excel Integration

Improved Performance of the Excel

Reader node (XLS Reader).

Preview and Execution are now much

faster.

XLS Writer Node

Multi-sheet

Writing

Open xls

File after

Execution

Column

Selection

XLS Writer: Image Writing

New XLS Nodes

XLS Sheet Appender

Appends a new

spreadsheet to an

existing excel file

Read XLS

Sheet Names

Provides the

spreadsheet names

of an excel file at

the output port

Database Nodes

Time

Zone

Default JDBC

Drivers

Zip/Unzip Nodes

Zips a list of files,

whose paths are taken

from a data column

Unzip files from an

archive and reports

the file paths at the

output port

OSM Integration

OSM Nodes visualize

points on a map based

on the longitude and

latitude coordinates

The 50 largest cities

of the world

The New Math Formula Node

Similar to the String

Manipulation Node

Now flow variable

values can be

inserted in math

formula

Error Handling

Default Data available when main branch fails

(ex: an error message)

Lag Column Node (Lag)

x(t) , x(t-1), ..., x(t-Lag)

x(t-1), x(t-2), ..., x(t-Lag-1)

x(t-2), x(t-3), ..., x(t-Lag-2)

x(t) x(t-1) x(t-2) … x(t-Lag) x(t-n)

Example: Values for the whole past week

(Lag=7 on daily values)

Lag Column Node (Lag Interval)

x(t), x(t - Lag Interval)

x(t-1), x(t – Lag Interval – 1)

x(t-2), x(t – Lag Interval – 2)

x(t) x(t-1) x(t-P) x(t-n)

Example: Values from every Monday

(Lag Interval=7 on daily values)

Lag Column Node (Lag * Lag Interval)

x(t), x(t - Lag Interval), x(t – 2* Lag Interval), …, x(t – Lag*Lag Interval)

x(t-1), x(t - Lag Interval-1), x(t – 2* Lag Interval-1), …, x(t – Lag*Lag Interval-1)

x(t) x(t-1) x(t-P) x(t-2P)

Example: Values from every Monday for last 4

weeks (Lag Interval=7, Lag=4)

Lag Column Node

The Lag Column Node opens the way to time

series analysis.

Lag Interval = 7 * 24 hours

For seasonality correctionLag = n

R

R is a very good data analytics

software:

• Very extensive and impressive qulity

graphic library

• Cutting edge algorithms

But …

• It is a pain to learn and debug …

Improved R Integration

• New Interactive R Editor for easier

debugging

• New R(Interactive) category in KNIME

Labs with Interactive R Editor

• New R Source node

Interactive R Editor

R Variables

Inspection

R script

evaluation

results

Display

temporary

plot

Syntax

Highlight

Create an R

script template

Temporary Plot

Reuse an R

script template

New R(Interactive) category

New R Source Node

This node just reads data.

Library(‚foreign‘) and

library(‚Hmisc‘) can import

all data analyitcs formats:

SAS, SPSS, Stata, etc ...

Other Improvements

• PMML Ensemble Learning

• Text Processing

• Network Mining

• New QuickForms with new GUIs

• Additional IO nodes

• …

http://tech.knime.org/whats-new-in-knime-28