Integrating Darknet YOLOv3 Into Apache NiFi Workflows Darknet has released a new version of YOLO, version 3. This one is a faster and perhaps more accurate. It's new and shiny and I had to try it. I am liking the results. Flow to Execute Script We call the shell script, then I route out the empty results. I use SplitText to split into individual lines. I use Extract Text with ([^:]+):(.*) to split into our name, value pairs. We also want to process the images produced by YOLOv3. We grab the newest ones from the output directory. I also add meta data extraction and Tensorflow analysis. This data is stored in attributes and can be saved independently via using AttributesToJSON to build a new flow file that we save off separate probably converting into Apache ORC and storing in HDFS for Apache Hive querying. The image file we can store in the cloud, another file system, send to a front end or save in HDFS. Or even email it to someone. The parsed YOLOv3 results in Apache NiFi Attributes. As you can see we would grab labelvalue.1 and labelvalue.2 to do our processing. We may want to send this to JMS or MQTT or Apache Kafka for further display in an application or dashboard. This is an example of the result of our Extract Text. This is the output that we parse with Apache NiFi YOLOv3 also generates an image with rectangles and labels. YOLOv3 does some great classification on multiple items in a picture. I use Python to capture an image from my webcam via OpenCV2. I wrap my call in a shell script that captures the image sends it to Darknet's build of YOLOv3 and send errors to /dev/null. If you have a good GPU, you can compile with CUDA and OPENCV to do real-time off a webcam Example Output: /Volumes/seagate/models/darknet-master/images/yolo_image_img_20180514183707.jpg: Predicted in 26.351510 seconds. cell phone: 72% chair: 78% chair: 72% chair: 59% person: 100% chair: 83% Example Run: ./darknet detect cfg/yolov3.cfg cfg/yolov3.weights /Volumes/seagate/StrataNYC2018/kafka.jpg Source: https://github.com/tspannhw/nifi-yolo3/tree/master Reference: See: https://github.com/pjreddie/darknet See: https://pjreddie.com/darknet/yolo/ Download the training weights and data (https://pjreddie.com/media/files/yolov3.weights) See: https://pjreddie.com/media/files/papers/YOLOv3.pd @article{yolov3, title={YOLOv3: An Incremental Improvement}, author={Redmon, Joseph and Farhadi, Ali}, journal = {arXiv}, year={2018} } ... View more

Ethereum: Accessing Feeds from Etherscan on Volume, Price and Transaction (Apache NiFi) - Cryptocurrency Part 2 As opposed to EtherDelta that we accessed in our first post (https://community.hortonworks.com/content/kbentry/191146/accessing-feeds-from-etherdelta-on-trades-funds-bu.html) Etherscan has a familiar SSL HTTP REST call returning clean JSON. We are reading three REST APIs for acquiring three sets of interesting Ethereum data including logs, current supply and current price. Overview of Flow REST API Ingest feed for getLogs We split into individual JSON records for processing. Ether Supply and Ether Last Price Flows For these two flows they return very minimal amounts of data with some flags I don't care about (status, message). So I use EvaluateJsonPath to extract just the good fields. Then with UpdateAttribute I add a just created time stamp (now()). Finally I use AttributesToJSON to replace the existing flow file with a new JSON file built from the fields I want only. You can choose to include the system generated metadata attributes automagically, but I turned that off. Sometimes it's nice to have the uuid, filename and other system metadata for tracking. You can also send that data somewhere else via built-in Apache NiFi reporting tools. Our Save Everything Flow We added these three feeds: I added three new feeds to our save everything as Apache ORC files in HDFS flow from the EtherDelta ingest. You can see them on the right in Orange, Indigo and Violet. A rainbow of cryptocurrency storage, I am not sure what colors I will add next for additional feeds. The process works the same as it did for the other data, set a schema name and storage location in HDFS, QueryRecord processor to limit the data and convert to Apache AVRO and then finally merge these smaller Apache AVRO files into one bigger file (HDFS likes fewer, bigger files. In Hadoop 3.1, this is not as big a deal). Final step is to save in HDFS at the location field passed in. This last step in the Data Provenance will have the hive.ddl and HDFS location to build an external Apache Hive table. We can copy that and do that in Apache Zeppelin or you can hook up a PutHiveQL processor and have Apache NiFi build the table for you. I do this in Apache Zeppelin to control table builds. I also add mine to version control. To build this schema, I used the InferAvroSchema earlier then removed it from the flow. I paste the generated schema I grabbed from the schema attribute in Data Provenance into a new schema in Hortonworks Schema Registry, seen below. The schema registry allows for versioning and comparing versions. It's very useful and required for doing record processing in Apache NiFi, Hortonworks Streaming Analytics Manager and other tools. It's also very helpful for Apache Kafka. An example of our newly ingested supply data: An example of a QueryRecord Apache Calcite query: To learn more about querying with this awesome tool, see: http://calcite.apache.org/docs/ Apache Zeppelin Used to Create Our Tables Apache Zeppelin Queries of the Data Before we store data in HDFS, we have to build tables and possibly setup security with either hdfs dfs chmod or with Apache Range. HDFS Build Our Directories for Apache ORC File Storage hdfs dfs -mkdir -p /ethereum/tx hdfs dfs -mkdir -p /ethereum/supply hdfs dfs -mkdir -p /ethereum/price Create Our Tables in Hive View 2 or Beeline or Apache Zeppelin CREATE EXTERNAL TABLE IF NOT EXISTS ethereumprice (ethusd STRING, ethbtc STRING, ethusd_timestamp STRING, ethbtc_timestamp STRING, captureDate STRING) STORED AS ORC LOCATION '/ethereum/price'; CREATE EXTERNAL TABLE IF NOT EXISTS ethereumtx (address STRING, topics ARRAY<STRING>, data STRING, blockNumber STRING, `timeStamp` STRING, gasPrice STRING, gasUsed STRING, logIndex STRING, transactionHash STRING, transactionIndex STRING) STORED AS ORC LOCATION '/ethereum/tx'; CREATE EXTERNAL TABLE IF NOT EXISTS ethereumsupply (totalEther STRING, captureDate STRING) STORED AS ORC LOCATION '/ethereum/supply'; SQL Queries to Check out the Data from Apache Zeppelin select * from ethereumsupply order by capturedate desc; select max(ethusd) from ethereumprice p; select ethusd, capturedate from ethereumprice order by ethusd_timestamp desc limit 500; select * from ethereumtx order by `timeStamp` desc; select * from ethereumsupply order by capturedate desc; getLogs API (Transactions) from EtherScan https://api.etherscan.io/api?module=logs&action=getLogs &fromBlock=379224 &toBlock=latest &address=0x33990122638b9132ca29c723bdf037f1a891a70c &topic0=0xf63780e752c6a54a94fc52715dbc5518a3b4c3c2833d301a204226548a2a8545 &apikey=YourApiKeyToken Ethereum Price from EtherScan https://api.etherscan.io/api?module=stats&action=ethprice&apikey=YourAPIKeyToken Ethereum Volume / Supply from EtherScan https://api.etherscan.io/api?module=stats&action=ethsupply&apikey=YourApiKeyToken Raw Data Parsed Into Fields By JSON Path $.status $.message $.result / $.result.ethbtc $.result.ethbtc_timestamp $.result.ethusd $.result.ethusd_timestamp Apache Calcite / Apache NiFi - QueryRecord SQL SELECT captureDate, (CAST(totalEther AS DOUBLE)/1000000000000000000) as totalEther FROM FLOWFILE WHERE CAST(totalEther AS DOUBLE) > 0 Build a File Name ${filename:append('ethprice.'):append(${now():format('yyyymmddHHMMSS'):append(${md5}):append('.json')})} References https://etherscan.io/apis#logs https://etherscan.io/ https://www.ethereum.org/ https://etherscan.io/apis#stats https://github.com/ethereum/wiki/wiki/JSON-RPC#eth_getlogs https://calcite.apache.org/docs/reference.html NiFi Flow Template: allet.xml ... View more

Accessing Feeds from EtherDelta on Trades, Funds, Buys and Sells (Cryptocurrency Analysis) EtherDelta lets you trade Ether or Ethereum-based tokens. Ethereum (https://www.ethereum.org/) is an open-source block chain platform for running smart contracts. They provide a fast web-socket feed of all data coming through the system. We can tap this websocket feed with Apache NiFi to examine and ingest all the trades, funds, buys and sells coming through the system as JSON. Once we ingest, clean up, parse and schematize the data we can run queries on it with Apache Spark SQL and Apache Hive in Apache Zeppelin notebooks. Now the data scientists have a continuing stream of data to play with. Next we will start adding additional feeds from other Ethereum exchanges, Bit Coin APIs and other sources of data. Some of the APIs are REST, some are websockets and some are SDKs. All of these are easy to ingest with Apache NiFi. Initial Ingest from EtherDelta via Web Sockets API Route To The Correct Type To process this feed with Apache NiFi 1.6+: ConnectWebSocket - wss://socket.etherdelta.com/socket.io/?transport=websocket ReplaceText - three to remove extraneous data outside of the JSON (40, 42, 0 - websocket junk text) RouteOnAttribute - filter out too small files RouteOnContent - sell, buy, trades and funds SplitJson buys arrays ($.*.buys) or ($.*.sells) SplitJson buys and sells into individual JSON Records ($.[*]) UpdateAttribute add a schema name Send to Remote Cluster via HTTP/HTTPS On Remote Cluster process and query to limit and convert data to Hive friendly format Store in HDFS NiFi generates DDL for an external Hive table Query with Zeppelin Hand off to other Zeppelin user who is a Data Scientist for machine learning and statistics. Profit! Break apart Sell Records Setup Our Web Socket Client Macintosh Default Java Certificate for SSL (default password is changeit) Connect to the Web Socket Remove the extra '40' Throw away small files! Route for Buys, Funds, Sells and Trades EtherDelta Exchange https://etherdelta.com/#PPT-ETH https://github.com/etherdelta/bots https://github.com/etherdelta/etherdelta.github.io/blob/master/docs/API.md EtherDelta provides a WebSocket feed of their data, so I am ingesting that with Apache NiFi and breaking out different types of data being published. This will let us ingest into different Apache Hive tables and run some queries and analytics in Apache Zeppelin on this data. We can then make it available to Data Scientists. JsonPath Expressions JsonPath Expression for trades and funds $.*.* Split Json Orders Buys $.*.buys Split Json Orders Sells $.*.sells Schemas ethereumfunds { "type" : "record", "name" : "ethereumfunds", "fields" : [ { "name" : "txHash", "type" : "string", "doc" : "Type inferred from '\"0xc40e227f3d5c2e125791cd865e8dd36c4a1a86538f13905ce91d65d2ac721742\"'" }, { "name" : "date", "type" : "string", "doc" : "Type inferred from '\"2018-05-10T15:52:19.000Z\"'" }, { "name" : "tokenAddr", "type" : "string", "doc" : "Type inferred from '\"0xfb1e5f5e984c28ad7e228cdaa1f8a0919bb6a09b\"'" }, { "name" : "kind", "type" : "string", "doc" : "Type inferred from '\"Deposit\"'" }, { "name" : "user", "type" : "string", "doc" : "Type inferred from '\"0xb975cf6c40f9cc5ad5cb7a335f16bdaab6cdcf0d\"'" }, { "name" : "amount", "type" : "string", "doc" : "Type inferred from '\"5986.826\"'" }, { "name" : "balance", "type" : "string", "doc" : "Type inferred from '\"5986.826\"'" } ] } ethereumtrades { "type" : "record", "name" : "ethereumtrades", "fields" : [ { "name" : "txHash", "type" : "string", "doc" : "Type inferred from '\"0x0350b7b479c9372c07188d69aa642ced7637b05444735653d61316dd852a673c\"'" }, { "name" : "date", "type" : "string", "doc" : "Type inferred from '\"2018-05-10T18:32:27.000Z\"'" }, { "name" : "price", "type" : "string", "doc" : "Type inferred from '\"0.000049874\"'" }, { "name" : "side", "type" : "string", "doc" : "Type inferred from '\"buy\"'" }, { "name" : "amount", "type" : "string", "doc" : "Type inferred from '\"8800\"'" }, { "name" : "amountBase", "type" : "string", "doc" : "Type inferred from '\"0.4388912\"'" }, { "name" : "buyer", "type" : "string", "doc" : "Type inferred from '\"0xd170db528cd2dd6ca67b0b2e3f7cd6e24942dba2\"'" }, { "name" : "seller", "type" : "string", "doc" : "Type inferred from '\"0xecfd625bfc433e8f6c8ce4abb92b9e8f1db3e401\"'" }, { "name" : "tokenAddr", "type" : "string", "doc" : "Type inferred from '\"0x6888a16ea9792c15a4dcf2f6c623d055c8ede792\"'" } ] } ethereumbuy { "type" : "record", "name" : "ethereumbuy", "fields" : [ { "name" : "id", "type" : "string", "doc" : "Type inferred from '\"9c946737c29c807255c3aac7334e182e375cc3a32684c66ccda03e9f5c52e47e_buy\"'" }, { "name" : "amount", "type" : "string", "doc" : "Type inferred from '\"6.85381e+21\"'" }, { "name" : "price", "type" : "string", "doc" : "Type inferred from '\"0.0002743\"'" }, { "name" : "tokenGet", "type" : "string", "doc" : "Type inferred from '\"0x99ea4db9ee77acd40b119bd1dc4e33e1c070b80d\"'" }, { "name" : "amountGet", "type" : "string", "doc" : "Type inferred from '\"6.85381e+21\"'" }, { "name" : "tokenGive", "type" : "string", "doc" : "Type inferred from '\"0x0000000000000000000000000000000000000000\"'" }, { "name" : "amountGive", "type" : "string", "doc" : "Type inferred from '\"1880000083000000000\"'" }, { "name" : "expires", "type" : "string", "doc" : "Type inferred from '\"5590043\"'" }, { "name" : "nonce", "type" : "string", "doc" : "Type inferred from '\"7586490717308181\"'" }, { "name" : "v", "type" : "int", "doc" : "Type inferred from '28'" }, { "name" : "r", "type" : "string", "doc" : "Type inferred from '\"0xd04a8f0a1f86fe8e3bdefb717f1bc461cfdb998f705dff3a5fc5567d023ca116\"'" }, { "name" : "s", "type" : "string", "doc" : "Type inferred from '\"0x7a8b687a8b7faea852ac873c02428acd5c26457282a4105fe12b60776fd87d55\"'" }, { "name" : "user", "type" : "string", "doc" : "Type inferred from '\"0x0b419BCE1Cb87ADEa84A913Fa903593fB68D33B1\"'" }, { "name" : "updated", "type" : "string", "doc" : "Type inferred from '\"2018-05-10T15:41:10.058Z\"'" }, { "name" : "availableVolume", "type" : "string", "doc" : "Type inferred from '\"6.85381e+21\"'" }, { "name" : "ethAvailableVolume", "type" : "string", "doc" : "Type inferred from '\"6853.81\"'" }, { "name" : "availableVolumeBase", "type" : "string", "doc" : "Type inferred from '\"1880000083000000000\"'" }, { "name" : "ethAvailableVolumeBase", "type" : "string", "doc" : "Type inferred from '\"1.880000083\"'" }, { "name" : "amountFilled", "type" : "null", "doc" : "Type inferred from 'null'" } ] } ethereumsell { "type" : "record", "name" : "ethereumsell", "fields" : [ { "name" : "id", "type" : "string", "doc" : "Type inferred from '\"27b06f4b8caf4aaa6d05841f8daa077f5f2131145331489ae94febc5eddd8c56_sell\"'" }, { "name" : "deleted", "type" : "boolean", "doc" : "Type inferred from 'true'" }, { "name" : "amount", "type" : "string", "doc" : "Type inferred from '\"-1.1627879935162941e+21\"'" }, { "name" : "price", "type" : "string", "doc" : "Type inferred from '\"0.0004473\"'" }, { "name" : "tokenGet", "type" : "string", "doc" : "Type inferred from '\"0x0000000000000000000000000000000000000000\"'" }, { "name" : "amountGet", "type" : "string", "doc" : "Type inferred from '\"520115069499838335\"'" }, { "name" : "tokenGive", "type" : "string", "doc" : "Type inferred from '\"0xe3818504c1B32bF1557b16C238B2E01Fd3149C17\"'" }, { "name" : "amountGive", "type" : "string", "doc" : "Type inferred from '\"1162787993516294126989\"'" }, { "name" : "expires", "type" : "string", "doc" : "Type inferred from '\"5589988\"'" }, { "name" : "nonce", "type" : "string", "doc" : "Type inferred from '\"57125161\"'" }, { "name" : "v", "type" : "int", "doc" : "Type inferred from '28'" }, { "name" : "r", "type" : "string", "doc" : "Type inferred from '\"0xafd5497f6159ac6589fd1804d27fe05436ed13706e64002f0e82e93b471e1780\"'" }, { "name" : "s", "type" : "string", "doc" : "Type inferred from '\"0x76dc38410d35069d1a62c08a1976548fbb915d4c52d8c9789157149720b04a33\"'" }, { "name" : "user", "type" : "string", "doc" : "Type inferred from '\"0x7418b4B9327b2DD18AC90Ef2eF846b36F286adA4\"'" }, { "name" : "updated", "type" : "string", "doc" : "Type inferred from '\"2018-05-10T15:36:55.000Z\"'" }, { "name" : "availableVolume", "type" : "string", "doc" : "Type inferred from '\"1.16278799351629411637773277515287799587448e+21\"'" }, { "name" : "ethAvailableVolume", "type" : "string", "doc" : "Type inferred from '\"1162.7879935162941\"'" }, { "name" : "availableVolumeBase", "type" : "string", "doc" : "Type inferred from '\"520115069499838340\"'" }, { "name" : "ethAvailableVolumeBase", "type" : "string", "doc" : "Type inferred from '\"0.5201150694998383\"'" }, { "name" : "amountFilled", "type" : "null", "doc" : "Type inferred from 'null'" } ] } Example JSON Data {"id":"237ab7693be71d35783941da9686f340d32b6d1e7332eedd0636b3e7b3725b93_sell","deleted":true,"amount":"-6000000000000000000","price":"0.02","tokenGet":"0x0000000000000000000000000000000000000000","amountGet":"120000000000000000","tokenGive":"0x219218f117dc9348b358b8471c55a073e5e0da0b","amountGive":"6000000000000000000","expires":"5594927","nonce":"4013451455","v":27,"r":"0x3965d8a9b074c6dcf25ebe10d39833f5ec6aa2d892aaec057dffd4368cd39f46","s":"0x01820eb919fe725c142f560c501742a89d41ca86b5594fdedc94f0e6f91bc97f","user":"0x0d4F98cb588c18FCC2695e2341112f066A915f80","updated":"2018-05-11T13:58:48.064Z","availableVolume":"798131841361720650","ethAvailableVolume":"0.7981318413617207","availableVolumeBase":"15962636827234412","ethAvailableVolumeBase":"0.01596263682723441","amountFilled":null} { "type": "record", "name": "ethereumfunds", "fields": [ { "name": "txHash", "type": [ "string", "null" ] }, { "name": "date", "type": [ "string", "null" ], "doc": "Type inferred from '\"2018-05-10T15:52:19.000Z\"'" }, { "name": "tokenAddr", "type": [ "string", "null" ], "doc": "Type inferred from '\"0xfb1e5f5e984c28ad7e228cdaa1f8a0919bb6a09b\"'" }, { "name": "kind", "type": [ "string", "null" ], "doc": "Type inferred from '\"Deposit\"'" }, { "name": "user", "type": [ "string", "null" ], "doc": "Type inferred from '\"0xb975cf6c40f9cc5ad5cb7a335f16bdaab6cdcf0d\"'" }, { "name": "amount", "type": [ "string", "null" ], "doc": "Type inferred from '\"5986.826\"'" }, { "name": "balance", "type": [ "string", "null" ], "doc": "Type inferred from '\"5986.826\"'" } ] } SQL Table DDL %jdbc(hive) CREATE EXTERNAL TABLE IF NOT EXISTS ethereumsell (id STRING, deleted BOOLEAN, amount STRING, price STRING, tokenGet STRING, amountGet STRING, tokenGive STRING, amountGive STRING, expires STRING, nonce STRING, v INT, r STRING, s STRING, `user` STRING, updated STRING, availableVolume STRING, ethAvailableVolume STRING, availableVolumeBase STRING, ethAvailableVolumeBase STRING, amountFilled STRING) STORED AS ORC LOCATION '/etherdelta/sell' %sql select * from ethereumsell order by cast(price as double) desc %jdbc(hive) CREATE EXTERNAL TABLE IF NOT EXISTS ethereumtrades (txHash STRING, `date` STRING, price STRING, side STRING, amount STRING, amountBase STRING, buyer STRING, seller STRING, tokenAddr STRING) STORED AS ORC LOCATION '/etherdelta/trade' %jdbc(hive) CREATE EXTERNAL TABLE IF NOT EXISTS ethereumbuy (id STRING, amount STRING, price STRING, tokenGet STRING, amountGet STRING, tokenGive STRING, amountGive STRING, expires STRING, nonce STRING, v INT, r STRING, s STRING, `user` STRING, updated STRING, availableVolume STRING, ethAvailableVolume STRING, availableVolumeBase STRING, ethAvailableVolumeBase STRING, amountFilled STRING) STORED AS ORC LOCATION '/etherdelta/buy' SELECT * FROM ethereumbuy order by cast(price as double) desc We now have four tables full of different Ethereum Trades, Sells, Buys and Funds Schemas in Hortonworks Schema Registry Storing The Data Run SQL on the Flows Reference: https://community.hortonworks.com/articles/68378/nifi-websocket-support.html https://gist.github.com/ijokarumawak/60b9ab2038ef906731ebf4c0eee97176 http://ijokarumawak.github.io/nifi/2016/11/04/nifi-websocket/ https://www.ethereum.org/cli http://web3py.readthedocs.io/en/stable/ Source etherdelta-analysis-1.json etherdelta-storage.xml ... View more