Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
veganism.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
Veganism Social is a welcoming space on the internet for vegans to connect and engage with the broader decentralized social media community.

Administered by:

Server stats:

294
active users

veganism.social: AboutStatusPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.0-alpha.5+vegan

#timeseries

2 posts2 participants0 posts today
Public
Miki :rstats:

#30DayChartChallenge Día 11: Stripes! Mi versión: ¡El código de barras del pánico del mercado! 😱

Este gráfico muestra una línea de tiempo (1993-2025) donde cada raya vertical representa un día en que el VIX cerró ≥ 30 (¡alta tensión!).

El concepto clave aquí es el **Volatility Clustering**: la alta volatilidad no se distribuye uniformemente, ¡viene en rachas! Los densos grupos de rayas identifican visualmente las grandes crisis (Dot-com, GFC '08, Covid '20...). Los largos periodos en blanco son la calma relativa.

Es una forma directa de ver la *persistencia* y los *regímenes* de la volatilidad del mercado. ¡Olvida las medias simples, el estrés viene en oleadas! 🌊

🛠️ Hecho con #rstats, #ggplot2, #quantmod.
📂 Código/Repo: t.ly/-vd9u

Gráfico estilo código de barras en una línea de tiempo de 1993 a 2025, visualizando días de alta volatilidad del mercado. El eje X muestra los años. Rayas verticales finas de color morado marcan cada día en que el índice VIX cerró igual o superior a 30. Se observan densos clústeres de rayas alrededor de periodos de crisis como 2001-2002, 2008-2009 y 2020. El eje Y está oculto. Título: "Días de Alta Volatilidad del Mercado (VIX >= 30)". Fuente: Yahoo Finance.
ALT
#Day11#Stripes#dataviz
Public
Petrônio Cândido

Foi publicado o artigo "A Multi-Step Multivariate Fuzzy-based Time Series Forecasting on Internet of Things Data" no periódico IEEE Internet of Things Journal de autoria dos incansáveis Hugo Bitencourt, Patricia Lucas, Omid Orang, eu e Frederico Gadelha Guimarães. Tenho muito orgulho de fazer parte deste grupo de pessoas que amam e não se cansam de pesquisar. Que time!

O artigo está disponível no link doi.org/10.1109/JIOT.2025.3549 e todo feedback é bem vindo!

#newpaper#research#timeseries
Public
Joris Meys

Very interesting paper in #Nature showing the phenological shift of #plants due to #ClimateChange . They combined a number of datasets on the occurence of changes that indicate spring, like flowering or leaf out. I show two of the longest time series, the leaf-out dates in the UK recorded by the Marsham family, and the blooming of cherry trees in Japan, documented in diaries and chronicles.

nature.com/articles/s41558-022

Cutout of figure 1 from the paper. It shows the day of the year spring phenology is detected for the pedunculate oak in southern UK and the cherry tree in Japan. For the oak the day when the leaves come out is tracked through time, and for the cherry tree it's the day of blossoming. For the oak you see a dataset from the 18th century that shows on average day 110 as day all leaves are out. A second line starts at 1840 and goed until mid 20th century. Spring phenology starts between 100 and 110 days in that period. The third dataset shows this day dropping from about day 120 in 1950 to below 100 after 200. The cherry tree dataset starts in 1500 and is continuous. It stays largely stable around day 105. Around 1940 it starts to drop fast to reach day 90 after the year 2000.
ALT
#Science#climate#DataScience
Public
Thomas Arildsen

I have a #timeseries of values at low temporal resolution where the values represent an average of the respective surrounding intervals.
I wish to up-sample this sequence to a higher temporal resolution in such a way that the average of the up-sampled values is equal to the corresponding value from the original time series.
Does an #algorithm for the kind of interpolation I am looking for exist? (not Pandas' resample or SciPy's signal.resample.) And is there an implementation of it in #Python?

Public
💧🌏 Greg Cocks

Mapping Historical Forest Biomass For Stock-Change Assessments At Parcel To Landscape Scales
--
doi.org/10.1016/j.foreco.2023. <-- shared paper
--
phys.org/news/2022-06-historic <-- shared technical article
--
dx.doi.org/10.1126/science.abk <-- shared paper
--
#GIS #spatial #mapping #Landsat #LiDAR #AbovegroundBiomass #MachineLearning #LandCover #biomass #spatialanalysis #spatiotemporal #historical #parcel #landscape #Landtrendr #landsat #timeseries #machinelearning #algorithm #model #modeling #field #fieldwork #inventory #LiDAR #NewYork #forest #vegetation #carbonstocks #carbonaccounting #temporal #opendata #opensource #climatechange #planning #management

photo - base of a moss-covered tree in a forest
ALT
schematic - A flowchart diagram showing the key elements of the modeling and mapping methodology. Cylinders represent data repositories, parallelograms represent data products and results, rectangles represent processing steps, and ovals represent models.
ALT
maps - New York State (USA) AGB difference map (2019 AGB - 1990 AGB) with predictions from the ensemble model. a) Statewide scale. b) A mosaic of working forests and Adirondack Forest Preserve land south of Stillwater Reservoir, NYS. c) Allegany River area with a portion of Allegany State Park to the east of the river. d) Forest growth along the border of the Catskill Forest Preserve. Values are capped at ± 75 Mg ha−1 for display.
ALT
maps - A qualitative comparison of maps from each modeling approach within Huntington Wildlife Forest (boundary mapped with black box in NAIP panel) and the surrounding area in Newcomb, New York (full area extent mapped with black box in New York State panel). Pair-wise raster subtractions (values capped at ± 75 Mg ha−1 for display) highlight spatial patterns and magnitudes of differences between model predictions. Ensemble - Direct not shown because it duplicates Indirect - Ensemble. A 1 m LiDAR-derived canopy height model and 0.5 m natural color National Aerial Imagery Program (NAIP) orthophotography included for additional reference information. All surfaces represent conditions in 2015.
ALT
Public
💧🌏 Greg Cocks

Integrated Topographic Corrections Improve Forest Mapping Using Landsat Imagery
--
doi.org/10.1016/j.jag.2022.102 <-- shared 2022 paper
--
“HIGHLIGHTS:
• [They] evaluated the impacts of topographic correction on forest mapping in the mountains.
• The enhanced C-correction and the physical model reduced topographic effects.
• The corrected Landsat imagery time series resulted in higher accuracy.
• Terrain information improved classification but not as much as topographic correction.
• [They] recommend using topographic correction for forest cover mapping..."
#GIS #spatial #AtmosphericCorrection #IlluminationCondition #LandCover #ModelComparison #TimeSeries #TopographicCorrection #remotesensing #comparasion #topographic #correction #NDVI #forest #vegetation #model #modeling #spatialanalyis #accuracy #forestcover #Russia #Georgia #CaucasusMountains #spatiotemporal #landsat #elevation #DEM

graphic - Data processing workflow
ALT
photo - Mount Elbrus
ALT
maps / images - Landsat summer (panel A) and autumn (Panel B) images (RGB: 743) for the study area. Subset region marked in black frame. Panels C-F show uncorrected (panel C), corrected summer image using the enhanced C-correction (panel D) or the physical model (panel E), and the corresponding illumination condition (F). Panels G-J show the same as panel C-F, but for the autumn image. The QA layers generated from FORCE were not applied here.
ALT
map - Forest cover classification agreement among the 18 sets of input variables. Pixels in red color were classified by all sets of input variables as coniferous, in green color as broadleaf forest, and in blue color as mixed forest. Black color indicates that no forest was predicted by any set of input variables. Two subsets A and B which are marked in white frames were zoomed in for a detailed map comparison in [another figure]
ALT
Public
Rami Krispin :unverified:

(1/3) I am excited to run a workshop tomorrow about analyzing time series with cluster analysis methods as part of the "Workshops for Ukraine" series organized by Dariia Mykhailyshyna. 

📆 When: Thursday, October 17th, 18:00 - 20:00 CEST

Registration required a donation of at least 20 EUR.

GIF
#AI#DataScience#Data
Replied to Estelle Platini
Public
Estelle Platini

"The ‘science of finance’ is first and foremost a collective ethos. Its real achievement is not objective discovery but ethical articulation. […] It fixes the underlying terrain, it shows them the proper path to follow, and it compels them to stay on track. Without this anchor, all capitalists — whether they are small, anonymous day traders, legendary investors such as Warren Buffet, or professional fund managers like Bill Gross — would be utterly lost.

"Finance theory establishes the elementary particles of capitalization and the boundaries of accumulation. It gives capitalists the basic building blocks of investment; it tells them how to quantify these entities as numerical ‘variables’; and it provides them with a universal algorithm that reduces these variables into the single magnitude of present value."

(Nitzan and Bichler, 2009) 🧶

#ownership#economics#property
Public
Rami Krispin :unverified:

(1/3) Book of the Week - Forecasting: Principles and Practice ❤️❤️❤️

If I need to choose three books to take to a remote island, Forecasting: Principles and Practice by Rob J. Hyndman and George Athanasopoulos would be one of them (the ISLR and Statistical Rethinking would be the other ones 😎).

This book is one of the down-to-earth books for getting started with time series analysis and forecasting 🎯.

#Rstats#timeseries#forecasting
Public
Rami Krispin :unverified:

Another great talk from the PyData London 2024 - Backtesting and error metrics for modern time series forecasting by Kishan Manani

Backtesting is the time series forecasting equivalent to the machine learning cross-validation approach for training and testing models.

youtube.com/watch?v=dSTXd8Hx72

YouTubeKishan Manani- Backtesting and error metrics for modern time series forecasting | PyData London 2024By PyData
#Python#timeseries#forecasting
VegansExploreLive feeds
About