W35 Santa Margherita di Pula stats & predictions

Hoy en día, la tecnología ha transformado la forma en que seguimos el tenis. Desde aplicaciones móviles hasta análisis avanzados con inteligencia artificial, los fanáticos tienen acceso a información detallada que antes era inimaginable.

• Análisis Avanzado: Herramientas como Hawk-Eye proporcionan datos precisos sobre tiros y estrategias durante los partidos.
• Tecnología Wearable: Los dispositivos portátiles permiten monitorear el rendimiento físico de los jugadores en tiempo real.
• Virtual Reality (VR): Algunas plataformas ofrecen experiencias inmersivas que te permiten sentirte como si estuvieras dentro del campo.
• johnlewiswri/johnlewiswri.github.io<|file_sep|>/_posts/2019-06-05-constructing-a-river-network-in-r.markdown --- title: "Constructing a River Network in R" author: "John Lewis" date: "6/5/2019" output: html_document: toc: true toc_depth: '6' toc_float: collapsed: false number_sections: true theme: cerulean highlight: tango code_folding: show --- {r setup, include=FALSE} knitr::opts_chunk$set(echo = TRUE) ## Introduction The purpose of this document is to illustrate how to construct river networks using R by combining several data sources together into one network. This includes (1) the National Hydrography Dataset (NHD) PlusFlow files (for flowlines and their attributes), (2) NHDPlusV2 vector data (for attributes and flow accumulation), and (3) NHDPlusV2 raster data (for flow direction). The resulting network can be used for various applications such as water quality analysis and hydrologic modeling. ## Data ### Flowlines We will use the [National Hydrography Dataset PlusFlow](https://www.sciencebase.gov/catalog/item/56b36d27e4b0f7568b43cc7c) which is available on ScienceBase as an example for this tutorial. This is the version of the dataset that includes attributes such as stream order and drainage area from the [NHDPlusV2](https://www.sciencebase.gov/catalog/item/55c6f0d6e4b0bd0c5127e19d) database. First we will read in the Flowline shapefile: {r echo=TRUE} library(sf) # Read in flowline shapefile flowline <- st_read("data/NHDPlus/NHDPlus_H08_Features/NHDPlus_H08_Features.shp") Next we will check to see what the columns are: {r echo=TRUE} # Print column names names(flowline) Now let's take a look at some of the columns that are important for constructing our network: {r echo=TRUE} # Check important columns head(flowline[, c("COMID", "GNIS_NAME", "StreamOrde", "DrainSqKm")]) We can also check out the projection of our data by looking at the CRS: {r echo=TRUE} # Check CRS st_crs(flowline) ### Flow Accumulation To create our network we will need to identify which COMIDs are tributaries of other COMIDs by looking at flow accumulation values from NHDPlusV2. This is available as raster data in ScienceBase item [NHDPlusV2](https://www.sciencebase.gov/catalog/item/55c6f0d6e4b0bd0c5127e19d). We will use `terra` to read in our raster data: {r echo=TRUE} library(terra) # Read in flow accumulation raster faccum <- rast("data/NHDPlus/NHDPlus_H08_Features/FOR_00005000_nhdplusv21_flowaccum.tif") We can also check the projection of this dataset: {r echo=TRUE} # Check CRS crs(faccum) To find tributaries we need to compare flow accumulation values to COMID values within each pixel: {r echo=TRUE} # Find tributaries tributary <- fcum > cellStats(faccum == flowline$COMID, fun = max, na.rm = TRUE, index = TRUE) Now let's take a look at what this looks like: {r echo=TRUE} plot(tributary, main = "Tributaries") ### Flow Direction We will also need flow direction raster data from NHDPlusV2 which is available on ScienceBase item [NHDPlusV2](https://www.sciencebase.gov/catalog/item/55c6f0d6e4b0bd0c5127e19d). Again we will use `terra` to read in our raster data: {r echo=TRUE} # Read in flow direction raster fdirection <- rast("data/NHDPlus/NHDPlus_H08_Features/FOR_00005000_nhdplusv21_fdir.tif") And check the projection of this dataset: {r echo=TRUE} # Check CRS crs(fdirection) ## Constructing the Network Now that we have all of our data ready we can begin constructing our network by finding tributaries and linking them together. First let's extract all COMID values that have tributaries: {r echo=TRUE} # Extract all COMIDs with tributaries tributary_comids <- unique(flowline$COMID[tributary]) Then let's find the parent COMID for each of these COMIDs by looking at which COMID has the highest flow accumulation value within each pixel: {r echo=TRUE} # Find parent COMIDs for tributaries parent_comids <- cellStats(faccum == flowline$COMID, fun = max, na.rm = TRUE, index = TRUE)[tributary] Now we can combine these two vectors into one dataframe: {r echo=TRUE} # Combine into dataframe network <- data.frame(COMID = tributary_comids, ParentCOMID = parent_comids) Next we need to find upstream COMIDs for each tributary so that we can connect them together later on. To do this we will loop through each tributary and look at its upstream neighbors until we reach its parent COMID. Let's start by creating an empty list to hold all upstream COMIDs for each tributary: {r echo=TRUE} # Create empty list for upstream COMIDs upstream_comids <- vector(mode = "list", length = nrow(network)) Now let's loop through each tributary and find its upstream neighbors: {r echo=TRUE} # Loop through each tributary and find upstream neighbors for(i in seq_len(nrow(network))) { # Initialize list of upstream COMIDs for current tributary # with itself and its immediate upstream neighbors upstream_comids[[i]] <- c(network$COMID[i], cellStats(fdirection == network$COMID[i], fun = sum, na.rm = TRUE, index = TRUE)) # While there are still more upstream neighbors that # haven't been visited yet... while(any(upstream_comids[[i]] %in% setdiff(flowline$COMID[fdirection], network$ParentCOMID[i]))) { # Update list of upstream COMIDs with new ones found # and repeat process until only parent COMID is left upstream_comids[[i]] <- unique(c(upstream_comids[[i]], cellStats(fdirection == upstream_comids[[i]], fun = sum, na.rm = TRUE, index = TRUE))) } # Remove parent COMID from list of upstream COMIDs # since it is not actually upstream of current tributary upstream_comids[[i]] <- setdiff(upstream_comids[[i]], network$ParentCOMID[i]) # Sort list of upstream COMIDs so they are in order # from closest to furthest away from current tributary upstream_comids[[i]] <- sort(upstream_comids[[i]]) } Now that we have all of our upstream COMIDs let's add them to our network dataframe as another column: {r echo=TRUE} # Add upstream COMIDs to network dataframe network$UpstreamCOMIDs <- lapply(upstream_comids, function(x) paste(x, collapse = ",")) Let's take a look at what our network looks like now: {r echo=TRUE} head(network) Finally let's join our network dataframe back onto our original flowline shapefile so that we have all of our original attributes along with our new network information: {r echo=TRUE} # Join network dataframe onto flowline shapefile network_flowline <- left_join(flowline, network) And here is what our final product looks like: {r echo=TRUE} head(network_flowline[, c("COMID", "GNIS_NAME", "StreamOrde", "DrainSqKm", "ParentCOMID", "UpstreamCOMIDs")]) ## Summary In this tutorial we showed how to construct a river network using R by combining several data sources together into one network. This included using NHD PlusFlow files for flowlines and their attributes, NHDPlusV2 vector data for attributes and flow accumulation, and NHDPlusV2 raster data for flow direction. The resulting network can be used for various applications such as water quality analysis and hydrologic modeling.<|file_sep|># John Lewis' Personal Website This website is built using Jekyll. You can view it at https://johnlewiswri.github.io <|repo_name|>johnlewiswri/johnlewiswri.github.io<|file_sep|>/_posts/2021-01-12-the-best-way-to-manage-your-r-packages.markdown --- title: "The Best Way to Manage Your R Packages" author: "John Lewis" date: "1/12/2021" output: html_document: toc: true toc