投稿時間:2022-02-08 08:40:52 RSSフィード2022-02-08 08:00 分まとめ(43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
TECH	Engadget Japanese	電気自動車に使用される電池を搭載。安全性を第一に考えたポータブル電源「PhewMan1200」	https://japanese.engadget.com/phewman1200-222532129.html	電気自動車に使用される電池を搭載。	2022-02-07 22:25:32
IT	ITmedia 総合記事一覧	[ITmedia News] Microsoft、OfficeのVBAマクロをデフォルトブロックへ　悪用対策で	https://www.itmedia.co.jp/news/articles/2202/08/news070.html	itmedianewsmicrosoft	2022-02-08 07:40:00
TECH	Techable（テッカブル）	川崎重工とソニーの新会社、約500km離れたリモート環境からロボットの遠隔操作に成功	https://techable.jp/archives/173007	事業会社	2022-02-07 22:00:48
AWS	AWS	Rappi Improves Efficiency, Stability, and Speed with Amazon DocumentDB | Amazon Web Services	https://www.youtube.com/watch?v=fDcW_4HeeHc	Rappi Improves Efficiency Stability and Speed with Amazon DocumentDB Amazon Web ServicesFounded in Rappi is Colombia s first tech startup to be valued at over billion It offers on demand last mile delivery for almost anythingーfrom food and clothing to insurance and financial servicesーin nine Latin American countries After trying other database solutions Rappi turned to Amazon Web Services AWS addressing its performance efficiency and scalability problems with Amazon DocumentDB with MongoDB compatibility Read the Case Study Visit the Amazon DocumentDB webpage Subscribe More AWS videos More AWS events videos ABOUT AWSAmazon Web Services AWS is the world s most comprehensive and broadly adopted cloud platform offering over fully featured services from data centers globally Millions of customers ーincluding the fastest growing startups largest enterprises and leading government agencies ーare using AWS to lower costs become more agile and innovate faster AWS AmazonWebServices CloudComputing AmazonDocumentDB DocumentDatabase Rappi	2022-02-07 22:21:17
js	JavaScriptタグが付けられた新着投稿 - Qiita	HTMLのTableタグをJavaScriptの配列に変換する	https://qiita.com/hsgw28/items/14ae71ce116596bbf68f	HTMLのTableタグをJavaScriptの配列に変換するHTMLのTableをcsvに変換するために色々調べてたんですが、その過程のJavaScriptでTable要素を配列に変換する方法をまとめておきます。	2022-02-08 07:18:46
海外TECH	Ars Technica	Health sites let ads track visitors without telling them	https://arstechnica.com/?p=1832346	company	2022-02-07 22:23:55
海外TECH	MakeUseOf	The 10 Best Side Business Ideas for Full-Time Workers	https://www.makeuseof.com/best-side-business-ideas-full-time-workers/	business	2022-02-07 22:31:11
海外TECH	MakeUseOf	5 Ways to Make Your Computer Read Documents to You	https://www.makeuseof.com/tag/5-ways-to-make-your-windows-computer-speak-to-you/	computer	2022-02-07 22:15:12
海外TECH	DEV Community	Fast Multivalue Look-ups For Huge Data Sets	https://dev.to/oblalex/fast-multivalue-look-ups-for-huge-data-sets-4pja	Fast Multivalue Look ups For Huge Data Sets Table Of ContentsSynopsisRationaleContextThe RecipeCaveats Synopsis This article provides a recipe for building fast look up tables for huge CSV data sets in Python Look up tables are meant to be read only data structures used to search for a presence of a given key and to retrieve multiple values associated with existing keys Recipe is meant to be a rough approach in contrast to a precise algorithm Huge is relative and means significantly exceeding the amount of available RAM of a single machine but comfortably fitting its storage Fast is subjective for a given use case and means quick enough to be used in stream processing or row by row batch processing The goal is achieved by calculating numeric representations of target keys sorting and indexing the data set doing binary searches against indices and retrieving actual data using offsets Rationale The article itself is meant to be a memo of a generic approach which I would like to share with my colleagues and my future self Hence it intentionally does not provide many details However questions in the comments are welcome A hidden goal is to try dev to publishing platform as an alternative to Medium as the latter imposes certain difficulties for publishing technical articles Context The recipe describes the essence of a subtask of a production daily batch job The job was processing datasets consisting of parquet files having millions of rows and around columns each Among columns there were SKUs and image URIs Both strings All rows were grouped by SKUs hence image URIs were stored as arrays Arrays had up to items with log normal distribution having the center of mass between and items Additionally each dataset had multiple versions and a single SKU could have differences in URIs between versions The goal of the subtask was to validate and fix all URIs associated with each given SKU present in multiple versions of datasets URIs themselves were pointing to an internal storage where file names were grouped by SKUs Hence in order to validate and fix URIs in data a listing of actual files needed to be obtained and stored This was done by traversing the storage using several workers that produced CSVs Each CSV was a flat list of file names associated with SKUs When catenated into a single final file listings were several gigabytes in size So in order to complete the goal for each SKU from each dataset version each image from an array of SKU s URIs had to be checked against a flat listing of actual files Essentially an inner join needed to be performed between each version of a dataset and a file listing both having up to tens of millions of items Conceptually in an ideal world this could be achieved by loading a dataset and a file listing as pandas dataframes explode ing URIs array normalizing them joining dataframes and then grouping the resulting dataframe by SKUs in order to get back arrays The issue is that doing so required up to GiB of RAM And the job was running in a Kubernetes cluster with limited resources having only a couple of gigabytes of RAM available Going beyond that made the job just die silently Basically GiB of RAM consumption was not the limit as in the future both datasets and file listings could grow unpredictably Hence another approach was needed and the priorities were RAM consumption and implementation speed as the task was urgent The recipe below describes one such possible approach which proved to be quick to implement comfortable to debug smooth to run and possible to control It allowed reducing the overall RAM consumption of the whole job from GiB to GiB which is × The Recipe Create a numeric hash of a data key e g SKU make it the st column in the CSV dataset file The hash itself does not need to be secure the only criteria are speed and consistency Python s built in hash will not work as it depends on random module hence it has no consistency Consider using FNV and pyfasthash as Python s implementation Use external merge sort to sort the CSV by the previously created numeric key This will group all data by the target key Consider using csvsort as Python s implementation Create an index of the data set The index is a rectangular data structure with fields Numeric key Offset of the first row with that key in the sorted dataset file Number of rows associated with that key Use numpy array for that and store it as a file At the moment numpy allows saving only the whole array and there s no way to append data chunks to a previously saved file Hence just in the case the whole index cannot fit RAM at once use npy append array to build the index by chunks Create a small wrapper for data look ups It should know about the locations of the CSV file and the index file but it should encapsulate all internals from clients providing a simple API given an original key e g SKU return all values associated with it The wrapper should take care of opening and closing of both files consider implementing the context manager for it Use numpy load with mmap mode r to load the index this will ensure the index will fit RAM by using making the file memory mapped See numpy memmap for a detailed description of modes In order to search a given original key convert it to a numeric key the same way this was done at previous steps Then use numpy searchsorted to perform a binary search in the index If a key is found seek the CSV file at the offset and read the count number of rows Return them as a result in task specific format Caveats The implementation of pyfasthash uses a binary compiled module under the hood Currently it does not compile on arm chips because compilation flags are hardcoded This can be an issue for users of MacBooks with M chips If this is the case confider using py fnvhash c or falling back to pure Python implementation during development via py fnvhash Or try another library or another algorithm which will be simple enough and consistent csvsort is a bit stale and has several issues Its main branch contains several nice improvements but they were not released for a long time Also one may wish to tune it up a bit for example to change the type for numeric values from float to int Fortunately it is a simple enough single module library which can be imported into a project and modified However that might not be an option for everyone numpy searchsorted function accepts only D arrays but the proposed index is D so slicing needs to be done Also the function always returns a position of the key inside the given array The result is a number in range len both sides included There s no special number like to signal the item was not found So the result must be validated Check if the position is less than the length of the index Read the index at the position and get a key offset count tuple If the key really equals what was passed to searchsorted then it is the actual result If it differs then nothing was found	2022-02-07 22:23:31
海外TECH	DEV Community	Leakage on iPhone 14 features 🍏	https://dev.to/yokwejuste/leakage-on-iphone-14-features-4im7	Leakage on iPhone features IntroductioniPhone as a phone is one with so many updates and changes brought in every time Before the year and before the official release of the iPhone some leakages on information about iPhone got out of the company iPhone sizes are changing in and Apple is eliminating the inch iPhone mini as it turned out to be unpopular with customers After lacklustre sales Apple is planning to focus on larger iPhone sizes for its flagship devices and we re expecting to see a inch iPhone a inch iPhone Pro a inch iPhone Max and a inch iPhone Pro Max The UpdatesNo Notch and New Face ID Hardware DesignThe biggest change that we currently expect with the iPhone is the removal of the notch cutout The notch design was first introduced with the iPhone X and it has stayed the same since then The notch is where Apple houses its sensors for things like Face ID the punch hole design will only come to fruition if Apple s prototyping and production go according to plan Apple s current goal is to bring the hole punch design only to the high end iPhone Pro and iPhone Pro Max As such the standard iPhone will retain the notch design unless Apple exceeds product yield goals Body DesignAs concerned with the design the iPhone is expected to look like iPhone Some features like volume buttons of round shape and thickness unlike features like the size and quality of the material being made with will be updated There will be no inch iPhone in because Apple is doing away with the mini line following lacklustre sales The iPhone mini will be the last of the mini phones and going forward Apple is expected to focus on larger sized iPhones We re expecting a inch iPhone a inch iPhone Pro a inch iPhone Max and a inch iPhone Pro Max with the larger inch iPhone replacing the mini model No camera bumpiPhone will not feature a camera bump Currently the iPhone features a protruding camera bump because of the thin design of the iPhone and the necessary thickness of camera components Camera improvementsThe iPhone will also include improvements to the front facing camera according to Kuo The biggest change will be upgraded to the front facing camera s auto focus features More details here are unclear but the front facing iPhone camera has a way to go to catch up to the rear camera Car Crash DetectionApple is working on a crash detection feature for the iPhone and the Apple Watch which could come out in It will use sensors like the accelerometer to detect car accidents when they occur by measuring a spike in gravitational force When a car crash is detected the iPhone or the Apple Watch would automatically dial emergency services to get help Since it s planned for this could be a feature designed for the iPhone models and the Apple Watch Series though it s not likely to be limited to those devices It will be an expansion of the Fall Detection feature that s in existing Apple Watch and iPhone models No Physical SIM SlotThe iPhone models may launch without a physical SIM slot with Apple transitioning to an eSIM only design Apple is allegedly advising major U S carriers to prepare for the launch of eSIM only smartphones by which suggests some iPhone models may be the first to come without a SIM slot With the iPhone Apple provided some models without a nano SIM in the box with cellular plans instead of being able to be activated using eSIM An eSIM allows for cellular plans to be added to a device without using a nano SIM TB StorageOn the recent iPhone apple brought the memory to TB and a possibility of increasing to TB of memory space Bonus iPhone next release will have the ability to be folded ConclusionIn conclusion the official release date Apple is expected to introduce the iPhone models at an event that s likely to be held in September if Apple follows previous launch timelines	2022-02-07 22:14:43
海外TECH	Engadget	Google releases its last Pixel 3 security update	https://www.engadget.com/google-pixel-february-2022-update-security-223116754.html?src=rss	Google releases its last Pixel security updateDon t expect to receive more updates for your Pixel Esper s Mishaal Rahman has learned Google is delivering one last security update to the Pixel and XL The company previously said the Pixel would get one last update in the first quarter of and confirmed to Engadget this represents the device s last hurrah It s not certain what the patch fixes However DarkPlayer noted the build ID matches that for an October patch that targeted newer Pixel models on Verizon This is a catch up release rather than an up to the moment patch You can expect more if you have a recent Pixel at least Google has posted a February update for the Pixel a and newer phones The revision tackles several significant problems including reboots during camera use Bluetooth audio disconnection and quality problems carrier specific connection woes and keyboards that override input text in some cases This won t thrill you if you re fond of the Pixel ーyou ll have to upgrade if you want up to the minute security fixes Don t fret if you have a Pixel though Google has promised five years of security updates for its latest phone line so you might not have to worry about patches until The Pixel and XL are getting what s presumably their final updateNot sure what s included in the update Has anyone gotten the update on their Pixel or XL yet ーMishaal Rahman MishaalRahman February	2022-02-07 22:31:16
海外科学	NYT > Science	Water Supplies From Glaciers May Peak Sooner Than Anticipated	https://www.nytimes.com/2022/02/07/climate/glaciers-water-global-warming.html	glaciers	2022-02-07 22:58:30
海外科学	NYT > Science	Climate Change Enters the Therapy Room	https://www.nytimes.com/2022/02/06/health/climate-anxiety-therapy.html	skepticism	2022-02-07 22:16:04
金融	金融総合：経済レポート一覧	季節調整ガチャに攪乱された1月雇用統計 ただしＦＥＤがきれいなデータを待つとは考えにくい：Market Flash	http://www3.keizaireport.com/report.php/RID/483893/?rss	marketflash	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	FX Daily（2月4日）～ドル円、米雇用統計受け115円台前半に上昇	http://www3.keizaireport.com/report.php/RID/483895/?rss	fxdaily	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	世界の分断を象徴する北京五輪：世界経済・金融市場への影響は？：木内登英のGlobal Economy ＆ Policy Insight	http://www3.keizaireport.com/report.php/RID/483899/?rss	lobaleconomypolicyinsight	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	英国金融政策（2月ＭＰＣ）～追加利上げと保有資産の縮小を決定：経済・金融フラッシュ	http://www3.keizaireport.com/report.php/RID/483904/?rss	保有資産	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	IAIGsの指定の公表に関する最近の状況（５）～情報が更新され、新たに１グループが追加され50グループに：保険・年金フォーカス	http://www3.keizaireport.com/report.php/RID/483906/?rss	iaigs	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	日本企業の価格転嫁スタンスの明確化に見るインフレの芽：金融市場	http://www3.keizaireport.com/report.php/RID/483923/?rss	日本企業	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	「国家戦略」としてのFATF対応：金融市場	http://www3.keizaireport.com/report.php/RID/483924/?rss	野村総合研究所	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	バーゼル委員会：気候関連金融リスク管理に関する諸原則（市中協議文書）を公表：リスク管理	http://www3.keizaireport.com/report.php/RID/483925/?rss	野村総合研究所	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	参入が相次ぐ「ミニ保険」の今後の戦略：保険ビジネス	http://www3.keizaireport.com/report.php/RID/483926/?rss	野村総合研究所	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	国内NFTマーケットプレイスの現状とセキュリティ対策：デジタルイノベーション	http://www3.keizaireport.com/report.php/RID/483927/?rss	野村総合研究所	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	「責任ある投資家」として日系保険会社に求められる脱炭素対応：保険の2030年アジェンダ	http://www3.keizaireport.com/report.php/RID/483928/?rss	保険会社	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	auじぶん銀行様で、勘定系システム、フロントシステムを同時更改：Case Study	http://www3.keizaireport.com/report.php/RID/483929/?rss	casestudy	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	みずほ経済・金融ウィークリー 2022年2月7日号～先週の内外経済・金融市場動向・評価＆今週の注目点	http://www3.keizaireport.com/report.php/RID/483931/?rss	金融市場	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	ESGの経済的現実	http://www3.keizaireport.com/report.php/RID/483953/?rss	pwcjapan	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	令和3年版「犯罪収益移転危険度調査書」に係る考察	http://www3.keizaireport.com/report.php/RID/483954/?rss	pwcjapan	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	日銀YCCは「犬の躾」～日銀に躾けられていた市場は「野生の本性」を発揮するか：高田レポート	http://www3.keizaireport.com/report.php/RID/483959/?rss	岡三証券	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	強まるドルの先高観 / ユーロ:年内利上げが視野に / 英ポンド:利上げ期待根強いが失速リスクも：Weekly FX Market Focus	http://www3.keizaireport.com/report.php/RID/483969/?rss	weeklyfxmarketfocus	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	グローバル・リート・マンスリー 2022年2月号～先進国リートは大幅下落、米金融引き締め観測に伴う金利上昇が重石	http://www3.keizaireport.com/report.php/RID/483970/?rss	三菱ufj	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	投資環境ウィークリー 2022年2月7日号【日本、米国、欧州、オーストラリア】主要国中銀は景気を冷やすことなく金融緩和の修正を進められるか？	http://www3.keizaireport.com/report.php/RID/483971/?rss	三菱ufj	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	【投資環境レポート2月号】 投資の視点：2022年のリスクシナリオ	http://www3.keizaireport.com/report.php/RID/483972/?rss	野村アセットマネジメント	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	【注目検索キーワード】トランジション・ファイナンス	http://search.keizaireport.com/search.php/-/keyword=トランジション・ファイナンス/?rss	検索キーワード	2022-02-08 00:00:00
金融	金融総合：経済レポート一覧	【お薦め書籍】5秒でチェック、すぐに使える! 2行でわかるサクサク仕事ノート	https://www.amazon.co.jp/exec/obidos/ASIN/4046053631/keizaireport-22/	結集	2022-02-08 00:00:00
金融	ニュース - 保険市場TIMES	東京海上日動ら、「リスクマネジメント動向調査2021」結果を発表	https://www.hokende.com/news/blog/entry/2022/02/08/080000	東京海上日動ら、「リスクマネジメント動向調査」結果を発表上場企業・従業員数規模名以上の非上場企業を対象東京海上日動火災保険株式会社以下、東京海上日動と東京海上ディーアール株式会社以下、東京海上ディーアールは年月日、企業のリスクマネジメントの動向に関する調査結果を発表した。	2022-02-08 08:00:00
ニュース	BBC News - Home	Keir Starmer: Two arrested after protesters surround Labour leader	https://www.bbc.co.uk/news/uk-politics-60294483?at_medium=RSS&at_campaign=KARANGA	westminster	2022-02-07 22:21:41
ニュース	BBC News - Home	Ukraine crisis: Nord Stream 2 will end if Russia invades - Biden	https://www.bbc.co.uk/news/world-europe-60292437?at_medium=RSS&at_campaign=KARANGA	ukraine	2022-02-07 22:37:23
ニュース	BBC News - Home	Duchess of Cambridge to read CBeebies Bedtime Story	https://www.bbc.co.uk/news/entertainment-arts-60290064?at_medium=RSS&at_campaign=KARANGA	children	2022-02-07 22:30:10
ニュース	BBC News - Home	Manchester United boss Ralf Rangnick believes club getting better despite FA Cup exit	https://www.bbc.co.uk/sport/football/60295830?at_medium=RSS&at_campaign=KARANGA	Manchester United boss Ralf Rangnick believes club getting better despite FA Cup exitManchester United interim boss Ralf Rangnick says the club are improving on the pitch despite the shock FA Cup exit to Middlesbrough	2022-02-07 22:37:18
ビジネス	ダイヤモンド・オンライン - 新着記事	ウクライナ侵攻なら「ノルド・ストリーム2」停止＝バイデン氏 - WSJ発	https://diamond.jp/articles/-/295655	停止	2022-02-08 07:05:00
ニュース	THE BRIDGE	エンタメ特化メタバースで躍進、VTuberのバーチャルライブを支援する「VARK」【Monthly Pitch！注目スタートアップ】	https://thebridge.jp/2022/02/monthlypitch_new-startup-202202-vark	エンタメ特化メタバースで躍進、VTuberのバーチャルライブを支援する「VARK」【MonthlyPitch注目スタートアップ】本稿はベンチャーキャピタル、サイバーエージェント・キャピタルが運営するサイトに掲載された記事からの転載MonthlyPitch注目スタートアップではMonthlyPitch編集部と協力し、毎月開催されるピッチ登壇社から特に注目のスタートアップを毎週ご紹介していきます。	2022-02-07 22:15:32
ニュース	THE BRIDGE	Monthly Pitch！ 今月紹介の注目スタートアップ4社の顔ぶれ	https://thebridge.jp/2022/02/monthlypitch_new-startup-202202-cyberagentcapital-insight	MonthlyPitch今月紹介の注目スタートアップ社の顔ぶれ本稿はベンチャーキャピタル、サイバーエージェント・キャピタルが運営するサイトに掲載された記事からの転載MonthlyPitch注目スタートアップでは、MonthlyPitch編集部と協力し、毎月開催されるピッチ登壇社から特に注目のスタートアップを毎週ご紹介していきます。	2022-02-07 22:00:47