KEMBAR78
DevDay: Node Analytics with Python, Chainhaus | PPTX
Node Analytics with Python
jamiel@chainhaus.com
http://jamiel.io
Quick bio
• Details: http://Jamiel.io
• CEO, Chainhaus
• Education
• Advanced Corda Classes (Tokens, Oracles, RPC, Complex Flows / States)
• Dev
• Events
• Two products based on Corda – FloLedger & DesignedAssets
• Teach graduate blockchain, AI & machine learning
• Columbia Business School
• NYU
http://meetup.com/blockchainNYC
Key blockchain trajectory
• Interop
• Data
Predictive Analytics
• I’ve issued a Letter of Credit to party X, how long will it take for them
to sign?
• At what price should I price this asset given what I know about the
buyer?
• Can I detect anomalies in partner transactions that may indicate
fraud?
What tools?
• Tableau?
• Hadoop?
• Spark?
What do data scientists use?
• Java
• Kotlin
• Scala
What data scientists use
• R
• Python
Introducing PyCorda
• Open source
• Python
• PyCorda.com
Hello World- Obligation CorDapp
• PartyA
• Set H2 port to 55555
Read NodeInfos
import pycorda as pyc
partyA = pyc.Node(url, username, password)
partyA.get_node_infos()
Linear States
• View linear states table
• Issue an obligation
• View linear states table
Node Tables
Vault states
Key stores
Snapshots (dumps)
Search
Node REST API calls
Flexibility of DataFrames
• Can
• Filter
• Select
• Join
• Clean
• Append
• Concat
• Bulk modify
• Chart
• Etc.
Charting
Plotly
Driver Management
Installation – Requires Python 64-bit
Next version
• Log analysis
• Reg ex filters
• Support QueryableStates
• Deserializing states from DB into Python objects
• Database clean up (Transaction message table)
Thank you
jamiel@chainhaus.com
http://jamiel.io

DevDay: Node Analytics with Python, Chainhaus