投稿時間:2022-10-01 19:15:27 RSSフィード2022-10-01 19:00 分まとめ(14件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
python	Pythonタグが付けられた新着投稿 - Qiita	【Python】キーワード抽出　〜pkeを用いたキーフレーズ抽出から評価まで〜	https://qiita.com/mekadomio/items/ce1725f6b8c25b8daf64	amazon	2022-10-01 18:45:51
python	Pythonタグが付けられた新着投稿 - Qiita	Excelのカウント関数をPythonで実現するには（初心者向け）	https://qiita.com/pentyan0303/items/3530a61529c635d90089	excel	2022-10-01 18:26:39
Ruby	Rubyタグが付けられた新着投稿 - Qiita	pluckメソッドの使い方と注意点	https://qiita.com/kingdom0927/items/45e251b71e011ab3f8f7	pluck	2022-10-01 18:29:43
Ruby	Railsタグが付けられた新着投稿 - Qiita	pluckメソッドの使い方と注意点	https://qiita.com/kingdom0927/items/45e251b71e011ab3f8f7	pluck	2022-10-01 18:29:43
海外TECH	DEV Community	CDK AWS Cloudwatch Evidently Demo	https://dev.to/aws-builders/cdk-aws-cloudwatch-evidently-demo-2mo	CDK AWS Cloudwatch Evidently Demo AbstractCloudwatch evidently Safely launch new features and validates web application choices by conducting online experiments and then decide if your experiment should be terminated depending on the results of the experiment In this post we try the feature flag control for the user who performed the login process using Cognito It leverage the CDK typescript to provide cloudwatch evidently project feature launches and experiment also using AWS Amplify to provide web app and login method using Amazon cognito authentication and cognito identity pool Table Of ContentsOverview of Cloudwatch evidentlyCreate cloudwatch evidently project and its features launches and experimentsBuilding Evidently MetricsDeploy cloudwatch evidently stackUse Amplify to start webapp for testing evidentlyTest evidently featureConclusion Overview of Cloudwatch evidently You can use Amazon CloudWatch Evidently to safely validate new features by serving them to a specified percentage of your users while you roll out the feature You can monitor the performance of the new feature to help you decide when to ramp up traffic to your users This helps you reduce risk and identify unintended consequences before you fully launch the feature Evidently structureProject The logical object in Evidently that can contain features launches and experiments Use projects to group similar features together We can store evaluation events for longer term storage by using cloudwatch log or S bucketFeature represents a feature that you want to launch or that you want to test variations of Launch To expose a new feature or change to a specified percentage of your users create a launchExperiment Use experiments to test different versions of a feature or website and collect data from real user sessions This way you can make choices for your application based on evidence and data Create cloudwatch evidently project and its features launches and experiments First we create a project for the application feature and store evaluation events to S bucket For CDK at the time of writing this there s only L construct const s new Bucket this prefix evidently demo data storage bucketName prefix evidently demo data storage blockPublicAccess BlockPublicAccess BLOCK ALL encryption BucketEncryption S MANAGED removalPolicy RemovalPolicy DESTROY enforceSSL true const proj new CfnProject this prefix evidently demo description S bucket to store evidently project evaluation events name prefix evidently demo dataDelivery s bucketName s bucketName tags InsideTags evidently reg Create a feature for the project The feature use Variation type Boolean with variations Variation false and Variation true represent for enabling or disabling the feature or in another words display hide or on off feature on the application const feature new CfnFeature this prefix evaluation demo description Evaluation demo feature name prefix evaluation demo project proj name variations booleanValue false variationName Variation booleanValue true variationName Variation defaultVariation Variation evaluationStrategy ALL RULES tags InsideTags evidently reg feature node addDependency proj Create launch split Variation and traffic to and set start now and add the launch to the feature const launch new CfnLaunch this prefix launch test name prefix launch test project proj attrArn groups groupName test launch feature feature name variation Variation groupName test launch feature feature name variation Variation scheduledSplitsConfig groupWeights groupName test launch splitWeight groupName test launch splitWeight startTime new Date toISOString executionStatus status START tags InsideTags evidently reg launch node addDependency feature Building Evidently Metrics Metrics are defined by applying rules to data events We use putProjectEvents to sends performance events to Evidently These events can be used to evaluate a launch or an experiment Evidently collects experiment data and analyzes it by statistical methods and provides clear recommendations about which variations perform better To make a demo of this We create a new feature similar to above and create an experiment associated with the feautre The experiment includes following propertiesmetricGoalswith desiredChange set to INCREASE means that a variation with a higher number for this metric is performing better The goal references to the custom metric and the metric rule bases on the eventPattern which is sent from putProjectEvents The rule entityIdKey entityId valueKey details loadTime eventPattern entityId exists true details loadTime exists true The event payload sent from application in evaluation processconst data entityId user username details loadTime elapse const event data JSON stringify data timestamp new Date type aws evidently evaluation onlineAbConfig defines treatment with specified splitWeight total treatments must be treatments associates treatments defined above with the feature and according variantSource code const featureExp new CfnFeature this prefix evaluation exp description Evaluation demo feature exp name prefix evaluation exp project proj name variations booleanValue false variationName Variation booleanValue true variationName Variation defaultVariation Variation evaluationStrategy ALL RULES tags InsideTags evidently reg featureExp node addDependency proj const exp new CfnExperiment this prefix experiment name prefix experiment project proj name description Test experiment metricGoals desiredChange INCREASE entityIdKey entityId metricName load time in second eventPattern JSON stringify eventPattern valueKey details loadTime onlineAbConfig controlTreatmentName prefix experiment treatment treatmentWeights splitWeight treatment prefix experiment treatment splitWeight treatment prefix experiment treatment treatments treatmentName prefix experiment treatment feature featureExp name variation Variation treatmentName prefix experiment treatment feature featureExp name variation Variation runningStatus status START analysisCompleteTime T Z exp node addDependency featureExp Deploy cloudwatch evidently stack The source code is ready we now deploy the stack to create cloudwatch evidently project feature and start the launch cdk deploy CloudwatchEvidentlyStack profile mfa concurrency require approval neverCheck the project created which has features launch and experimentFeature demo rule orders lauches gt defaultFeature with experiment Use Amplify to start webapp for testing evidently AWS Amplify is a complete solution that lets frontend web and mobile developers easily build ship and host full stack applications on AWS with the flexibility to leverage the breadth of AWS services as use cases evolve No cloud expertise needed By using Amplify we can build the Reac application with authentication by using cognito userpool and policy access control to AWS cloudwatch evidently by cognito identity poolIn this post we don t use CDK to provide Amplify components completely but adding Authentication component through Amplify studio or you can use amplify cli First create Amplify console with following settingsService role which Amplify requires permissions to deploy backend resources with your front endGitHubSourceCodeProvider connect amplify to github repository aws cloudwatch evidently react Note that Amplify already supports GitHub App to authorize access to repositories for CI CD workflows with least privilege buildSpec defines backend and frontend buid phasesSource code amplify console tsDeploy cdk deploy AmplifyConsoleReactStack profile mfaAfter deploy done following steps to setup backend and deploy build FE appWe should start migration to Install and authorize GitHub AppLaunch Backend environments studio to add Authentication componentClone backend staging to prod and assocaite main branch with prod backendTrigger build FE with the main branch after the build we will have the URL to access react appLet s have a look at react app source code to see how we create Evidently client to send evaluateFeatureRequest to the project and putProjectEvents for custom metric as experiment Test evidently feature After creating account and login successfully we will see the following errorIt s due to the identity pool default Authenticated role which is assumed by cognito userpool does not have permission to work with Cloudwatch evidently we need to provide AmazonCloudWatchEvidentlyFullAccess and also S permision to push data event storage to S bucket To limit the application permission toward cloudwatch evidently read Actions resources and condition keys for Amazon CloudWatch Evidently to create proper policy for the roleSuccessful load with evaluation routed to Variation with value falseSwitch app to experiment feature Conclusion With CDK we can provision cloudwatch evidently using infrastructure as code and can update modify or create new project feature launch experiment though cdk pipelineWe make a demo using Amplify for react app easily and improve security using cognito userpool and cognito identity pool through access token and role attached In practice where we don t use amplify we can inherrit the flow of the above authentication as best practice References New Amazon CloudWatch Evidently Experiments and Feature Management New feature Support feature flags and A B testing Amazon CloudWatch Evidently was released so I tried it reinvent Vu Dao Follow AWSome Devops AWS Community Builder AWS SA ️CloudOpz ️ vumdao vumdao	2022-10-01 09:15:42
海外TECH	CodeProject Latest Articles	EspMon: A simple PC hardware monitor using a T-Display S3	https://www.codeproject.com/Articles/5343526/EspMon-A-simple-PC-hardware-monitor-using-a-T-Disp	little	2022-10-01 09:53:00
海外ニュース	Japan Times latest articles	Japan to extend oil subsidy program to next spring to curb price hikes	https://www.japantimes.co.jp/news/2022/10/01/business/japan-oil-subsidy-gasoline/	Japan to extend oil subsidy program to next spring to curb price hikesSubsidies to curb the impact of a surge in gasoline and other fuel prices began in late January to keep the average retail price of	2022-10-01 18:27:21
海外ニュース	Japan Times latest articles	Can’t go forward, won’t go back: Truss’s government off to disastrous start	https://www.japantimes.co.jp/news/2022/10/01/world/politics-diplomacy-world/liz-truss-disaster-start/	mentality	2022-10-01 18:07:25
海外ニュース	Japan Times latest articles	U.S. clinches record-extending 11th women’s basketball World Cup	https://www.japantimes.co.jp/sports/2022/10/01/basketball/us-china-womens-world-cup/	vegas	2022-10-01 18:37:47
ニュース	BBC News - Home	Former Iran detainee 'determined' to finish Marathon	https://www.bbc.co.uk/news/uk-england-london-63101153?at_medium=RSS&at_campaign=KARANGA	ashoori	2022-10-01 09:14:23
北海道	北海道新聞	明大と早大が先勝　東京六大学野球第４週	https://www.hokkaido-np.co.jp/article/739241/	東京六大学野球	2022-10-01 18:12:15
北海道	北海道新聞	申ジエ首位、吉田１打差　日本女子オープン第３日	https://www.hokkaido-np.co.jp/article/739251/	日本女子オープン	2022-10-01 18:10:00
北海道	北海道新聞	東京で新たに３８３４人感染　新型コロナ、１３人死亡	https://www.hokkaido-np.co.jp/article/739250/	新型コロナウイルス	2022-10-01 18:10:00
北海道	北海道新聞	プリプリ厚岸カキうまい　牡蠣まつり開幕	https://www.hokkaido-np.co.jp/article/739249/	開幕	2022-10-01 18:05:00