#timeseries
This is super useful ๐ Amazon Timestream now supports unloading data to Amazon S3 ๐ https://aws.amazon.com/about-aws/whats-new/2023/05/amazon-timestream-unloading-data-amazon-s3/ #AWS #Analytics #TimeSeries #Database
To everyone in #Tokyo (#Japan) and you love #TimeScaleDB or want to learn more about #timeseries join me at #JJUG #CCC Spring 2023 in Shinjuku
skforecast 0.8.0 is available! ๐
skforecast is a Python library that makes it easy to perform time series forecasting using an API similar to the one of scikit-learn.
The new release brings improved support for categorical features, expanded backtesting functionality, performance improvements, and a new logo.
Install it with `pip install "skforecast[sarimax,plotting]==0.8.0"`
And have a look at the documentation! https://skforecast.org/0.8.0/
binjr 3.12 is now available!
This is a pretty featureful release, whose highlight is a new indexing mode for log files that makes is easy to search for arbitrary strings of character without the need for wildcards while maintaining the same level of performance, even on very large files.
It also features several other smaller enhancements and many bug fixes: full release notes available at https://binjr.eu/download/CHANGELOG/#binjr-v3120
๐ข๐ฎ๐น Calling all #TimeSeries and #XAI experts! ๐ฐ๏ธ๐๐ Join us in Torino, Italy for #ECML #PKDD and #XAITS - a workshop on Explainable AI for Time Series data! ๐๐ค Submit your paper on explainable and interpretable time series modelling before June 12th! Don't miss out on this exciting opportunity to showcase your research and connect with other XAI-TS enthusiasts (like me and @jakobs!) ๐๐๐ฉโ๐ฌ๐งโ๐ฌ๐จโ๐ฌ
I had the great pleasure of having Professor Huyรชn Pham to talk about #generative #modeling for #timeseries via Schrodinger bridge. It is the topic of a paper written with Mohamed Hamdouche and Pierre Henry-Labordere.
You can watch the recording of his presentation here:
https://www.youtube.com/watch?v=4OQvmO_Oxmw
You can read the paper here:
What is a Time Series database and why should it matter to you? #Timeseries #Voltmetrix #Grafana #timescaledb #InfluxDB #questdb
https://voltmetrix.com/blog/what-is-time-series-database/
Just presented my #PostgreSQL #extension #pg_statviz, for time series #analysis and #visualization of #Postgres internal #statistics, at #PGDay #Chicago :postgresql:!
Slides available here:
๐
https://postgresql.us/events/pgdaychicago2023/schedule/session/1215-announcing-pg_statviz/
#timeseries #timeseriesdata #opensource #database #databases #stats #performance #event #events
I had the great pleasure of having Professor Huyรชn Pham to talk about #generative #modeling for #timeseries via Schrodinger bridge. It is the topic of a paper written with Mohamed Hamdouche and Pierre Henry-Labordere.
You can watch the recording of his presentation here:
https://www.youtube.com/watch?v=4OQvmO_Oxmw
You can read the paper here:
๐ฟ๐ Excited to announce our new course on Ecological Time-series analysis & forecasting in R with Dr Nicholas J Clark! ๐
Learn to wrangle, visualize, & analyze ecological data with {mvgam} & {brms} packages.
Check it out:https://physalia-courses.org/courses-workshops/time-series-in-r/
Very excited to share that I'll be speaking at #PGDay #Chicago :postgresql: in less than a week! I'll be announcing the new #PostgreSQL #extension #pg_statviz for time series #analysis and #visualization of #Postgres internal #statistics.
Bonus talk: "Don't Do This" - PostgreSQL bad practices and pitfalls!
๐
https://2023.pgdaychicago.org/
#timeseries #timeseriesdata #opensource #database #databases #stats #performance #event #events
Are you planning to go to #Kubecon in #Amsterdam? Stop by and say hi to the team. We going to have a booth and we love to have a talk with you about #TimeSeries and #InfluxDB. https://www.influxdata.com/join-us-at-kubecon/
There were 173 new R packages on CRAN in February ๐ฆ
Check out this "Top 40" list compiled by Joseph Rickert including packages for #machinelearning, #timeseries, and #dataviz.
https://rviews.rstudio.com/2023/03/28/february-2023-top-40-new-cran-packages/
This R code serves as a great introduction to using the Shiny package for creating web applications and highlights the importance of model selection in forecasting. Give it a try and most importantly steal this code and see what you can do with it!
Post: https://lnkd.in/ehebft76
#shiny #r #rstats #opensource #software #timeseries #forecast #data #shinyapplication #code #technology #innovation
New (short) pre-print.
For time-series classification problems, we show that rock-bottom-simple models (that ignore the time-ordering of data!) can perform surprisingly well.
Some important general lessons for formulating and interpreting models.
Work led by @trentlikesstats
#Midjourney generated #illustrations using prompts from #science paper abstracts.
1-disentangle the interacting drivers of solute loading to streams
2-Anomalous patterns occurring in hydrological time series data
3-Salt marshes
4-space-for-time substitution
#TimeSeries #SciComm #CriticalZone
Outtakes from content created during #AGU22.
Sometimes we want to look at a #bootstrapping #model and we can easily achieve this in #r even in #timeseries #modeling with the #forecast #package
In this post I used the AirPassengers dataset and used the auto.arima() function to bootstrap models, then I wanted to take a look at the residuals.
This is not comprehensive but sheds some light on how to do this outside of the #tidymodels #framework
Post: https://lnkd.in/gwjMfezH
InfluxDB runs on Debian ARM 64-bit, so that would probably work. However, how performant is it on a Zero W? Do I need a Zero 2 W? #RaspberryPi #InfluxDB #timeseries #microcontroller
I want to run a time-series database on a Raspberry Pi Zero W. What do you recommend? #RaspberryPi #database #timeseries #microcontroller
Are you familiar with recurrence analysis?
I recently learned about it and have some questions about its applications and limits.
(Is it used outside academia? If so, for what? What are its criticisms, if any?)
From MLSTjournal on Twitter: NOW >18K DOWNLOADS!๐ฅณ'#Deeptime: a #Python library for #machinelearning dynamical models from #timeseries..' by @FrankNoeBerlin @eigensteve @andreasmardt @brookehus @tmhmpl et al @FU_Berlin @ME_at_UW @uwengineering @phybiofunc https://hcvalidate.perfdrive.com/fb803c746e9148689b3984a31fccd902//?ssa=a38c78ee-0f5e-43b6-ac4e-ba670ab9c138&ssb=91496224166&ssc=https%3A%2F%2Fiopscience.iop.org%2Farticle%2F10.1088%2F2632-2153%2Fac3de0&ssi=c4ed85a1-8427-4f8b-95c9-4fbf5faf024f&ssk=support@shieldsquare.com&ssm=86549507084850606109626231296620&ssn=7dd23543f6a7398e589502acb0d35c8375872082235a-4359-4d9c-b288af&sso=38540b82-894b6fe105628bfcb97bbabec4260625f7625669d32c7a39&ssp=54348371401678588416167851259553931&ssq=05830533672977305967136729652245230597516&ssr=MjQuMjIyLjQ0LjExMA==&sst=PycURL/7.45.1 libcurl/7.87.0 OpenSSL/3.0.7 zlib/1.2.13 libssh2/1.10.0 nghttp2/1.51.0&ssv=&ssw=&ssx=W10= #compchem #AI #dynamics #HPC https://twitter.com/MLSTjournal/status/1634447113097551875/photo/1
For the moment, we recommend to use the user-contributed #gretl package "auto_arima" written by @atecon .
https://gretl.sourceforge.net/current_fnfiles/auto_arima.gfn
Just to inform you:
The next release will include native support for #AutoARIMA automatically determining the optimal ARIMA #timeseries model specification.
This is will be pretty fast, as it's implemented in C.
I wrote about the design choice of single tables (with all data) and multiple tables (split by dimension); a common question for #timeseries data.
While the blog post is best-practices for #TimeScaleDB, the general sentiment behind the question is bigger!
Amazon Timestream now supports batch loading data ๐ A scalable solution to migrate data from CSV files stored in S3 directly into Timestream https://aws.amazon.com/about-aws/whats-new/2023/03/amazon-timestream-supports-batch-loading-data/ #AWS #Analytics #Storage #TimeSeries
My #JFokus talk "PostgreSQL, The Time-Series Database You Actually Want" about #timeseries on #PostgreSQL (and why #TimescaleDB is the convenience you're actually looking for) is online โค๏ธ๐
Our new paper "Combining embeddings and fuzzy time series for high-dimensional time series forecasting in the internet of energy applications" was just published in Elsevier's Energy journal and it is available at https://doi.org/10.1016/j.energy.2023.127072
Feedback is always welcome!
๐ Super excited to share that only after 4 years of being #opensource, the #stumpy #python package for modern #timeseries analysis just eclipsed 5+ Million downloads/installs and over 2500 #github stars โญ!
๐ Thank you to all of our active users and amazing contributors for maintaining such an active community! โค๏ธ The future of #matrixprofile is bright โ
Did you miss last week's webinar?๐
Watch the recording now to start leveraging time-series data for your business success๐๐
In this webinar, youโll gain expert insights on time-series data analysis and learn the crucial data modeling decisions needed to implement time-series data in CrateDB ๐ฉ๐ปโ๐ป
https://crate.io/resources/webinars/lp-wb-time-series-data
#crateio #cratedb #db #database #webinar #recording #livedemo #data #database #datascience #timeseries #timeseriesdata #timeseriesanalysis #dataanalysis
Buitre, Zhang & Lin (2019) use #timeseries analysis of #Landsat imagery using four #landscapemetrics to examine the impacts of #tropicalcyclones in #mangroves in Coron & Eastern Samar #Philippines. #LoLManuscriptMonday https://bit.ly/Buitre_2019
Shout-out to Mary Joy Buitreโs affiliated organizations and collaborators for this #LoLManuscriptMonday feature: DOST-PCIEERD, and CUHK! #EOChat ๐๐ฐ๏ธ
Calibrate and Plot a Time Series with {healthyR.ts}
Post: https://lnkd.in/e-9ixyZi
#data #training #datascientist #testing #r #rstats #opensource #opensourcesoftware #technology #innovation #softwareengineering #software #timeseries #modeling #forecasting
Hi all, Iโm a researcher working on brain-computer interfaces at Snap Inc.
My goal is to prove that noninvasive #BCI and #neurofeedback can provide value outside of the lab.
Currently for me that includes (but does not limit to) applying #deeplearning methods to #timeseries data like #EEG, but also other sensors like #eyetracking or inertial sensors (#IMUs)โจ
โจAlso a #python geek. I maintain an M/EEG analysis toolbox
โก๏ธ https://github.com/nbara/python-meegkit
Excellent presentation by Max Mergenthaler from Nixtla at #PyDataNYC on bridging classical and deep learning methods for #timeseries forecasting in #python thanks to #numba
Dear #datafam,
does anyone have recent examples of good #interactive time-series #dataviz?
I'm thinking of something like a linechart, possibly updated on scroll?
Hi, I'm Simon! I'm Data Scientist and #DataEngineer specializing in #remotesensing, #GIS, #spatial, #ecommerce, #timeseries, and duo of #Python + #opensource :)
Lived in ๐ซ๐ฎ , in ๐ต๐ฑ for 2023, then want to move further west ๐
Creator of #pyinterpolate package: https://pypi.org/project/pyinterpolate/ for Kriging and spatial interpolation ๐บ๏ธ
and #wsknn package for session-based recommendations: https://pypi.org/project/wsknn/ ๐
#Introduction Hi y'all! I'm not a newbie to tech in general, but I often feel like a newbie due to the breadth and depth of technology, especially software (and infrastructure)!
I sort of fell into IT through a random job that promoted me based on a "logic" test. :D
I have experience in #C, #COBOL, some #ruby, a tiny old crumb of #java, but most recently #go / #golang. *NIX wins, imho. PC for home, Mac for work. Spent some time in the #cloud. The team I'm on for work is in a decent-sized retailer, and our focus is Observability. #influx #elasticsearch #grafana #timeseries and some cool home-grown products. We support internal customers (other teams and platforms), and I love answering questions!
I'm married to my lovely wife, and our babies all have fur. I live in a brick house built in 1888. I love old houses, even though there's not a right angle or standard size left!
If it's horror, I love it - movies, books, objects. I'm a little #goth at heart (though I'm an Old Bat now).
GU #hoya (BSLA Russian Language). Breast cancer survivor. Sardonic and frequently inappropriate sense of humor. ๐
"With 36M+ Downloads, VictoriaMetrics Skyrockets to New Heights!" Read our 1st blog post on the new Vicky website, which provides a summary of our 2021 Momentum Milestones and more! With thanks to everyone who's contributed to last year's success:
bit.ly/334EqVF
#opensource #monitoring #timeseries #database #success