时间戳

首页 > 解决方案 > 按组对每一行和后续行应用一个函数

r - 按组对每一行和后续行应用一个函数

问题描述

我有一个 sf 数据框，其中包含标记许多单向街道交叉口位置的点。除了几何列之外，一列包含街道名称，另一列包含交叉点在单向街道上的相对位置。

下面是一个玩具示例。第一排是拱街的第一个路口，第二排是拱街的第二个路口，以此类推。

library(sf)

intersections <- structure(list(street = c("ARCH ST", "ARCH ST", "ARCH ST", "SANSOM ST", 
"SANSOM ST", "SANSOM ST"), number = c(1L, 2L, 3L, 1L, 2L, 3L), 
    geometry = structure(list(structure(c(2699665.2606043, 236074.947200272
    ), class = c("XY", "POINT", "sfg")), structure(c(2699402.74765515, 
    236109.729280198), class = c("XY", "POINT", "sfg")), structure(c(2699202.95996668, 
    236136.613760229), class = c("XY", "POINT", "sfg")), structure(c(2699431.38476158, 
    234437.663731016), class = c("XY", "POINT", "sfg")), structure(c(2699162.09261096, 
    234476.514355583), class = c("XY", "POINT", "sfg")), structure(c(2697100.77148795, 
    234809.605567052), class = c("XY", "POINT", "sfg"))), precision = 0, bbox = structure(c(xmin = 2697100.77148795, 
    ymin = 234437.663731016, xmax = 2699665.2606043, ymax = 236136.613760229
    ), class = "bbox"), crs = structure(list(epsg = 2272L, proj4string = "+proj=lcc +lat_1=40.96666666666667 +lat_2=39.93333333333333 +lat_0=39.33333333333334 +lon_0=-77.75 +x_0=600000 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=us-ft +no_defs"), class = "crs"), n_empty = 0L, class = c("sfc_POINT", 
    "sfc"))), row.names = c(NA, -6L), class = c("sf", "tbl_df", 
"tbl", "data.frame"), sf_column = "geometry", agr = structure(c(street = NA_integer_, 
number = NA_integer_), class = "factor", .Label = c("constant", 
"aggregate", "identity")))

> intersections
Simple feature collection with 6 features and 2 fields
geometry type:  POINT
dimension:      XY
bbox:           xmin: 2697101 ymin: 234437.7 xmax: 2699665 ymax: 236136.6
epsg (SRID):    2272
proj4string:    +proj=lcc +lat_1=40.96666666666667 +lat_2=39.93333333333333 +lat_0=39.33333333333334 +lon_0=-77.75 +x_0=600000 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=us-ft +no_defs

# A tibble: 6 x 3
  street    number                 geometry
  <chr>      <int> <POINT [US_survey_foot]>
1 ARCH ST        1       (2699665 236074.9)
2 ARCH ST        2       (2699403 236109.7)
3 ARCH ST        3       (2699203 236136.6)
4 SANSOM ST      1       (2699431 234437.7)
5 SANSOM ST      2       (2699162 234476.5)
6 SANSOM ST      3       (2697101 234809.6)

使用包中的mp_matrix()and ，我想添加一个列，显示从每个十字路口到该街道上的下一个十字路口的行驶时间（最后一个十字路口除外，它获得 NA）。mp_get_matrix()mapsapi

理想情况下，它如下所示：

     street number travel_time_sec                 geometry
1   ARCH ST      1             210 POINT (2699665 236074.9)
2   ARCH ST      2             180 POINT (2699403 236109.7)
3   ARCH ST      3              NA POINT (2699203 236136.6)
4 SANSOM ST      1             150 POINT (2699431 234437.7)
5 SANSOM ST      2             175 POINT (2699162 234476.5)
6 SANSOM ST      3              NA POINT (2697101 234809.6)

如何按组（即街道）循环遍历 sf 数据帧中的行，告诉每一行对该组中的下一行执行操作以填充新列，如果不存在这样的下一行，则返回 NA？

最后，由于mp_matrix()调用 Google Maps API 需要花钱，请改用st_distance()from 函数sf生成以下内容。

     street number travel_distance                 geometry
1   ARCH ST      1             576 POINT (2699665 236074.9)
2   ARCH ST      2             397 POINT (2699403 236109.7)
3   ARCH ST      3              NA POINT (2699203 236136.6)
4 SANSOM ST      1             410 POINT (2699431 234437.7)
5 SANSOM ST      2             440 POINT (2699162 234476.5)
6 SANSOM ST      3              NA POINT (2697101 234809.6)

非常感谢您的帮助。

标签： rsf

解决方案

我正在玩您的示例，但无法travel distance使用该st_distance功能。

st_distance(intersections$geometry[1], intersections$geometry[2])

Units: [US_survey_foot]
         [,1]
[1,] 264.8072

通过行本身的循环或矢量化操作可以用这段代码完成

# used librarys
library(units)
library(tidyverse)
library(sf)

# find distance function
find_Distance <- function(x) {

  # create lead list
  x_lead <- x[2:length(x)]

  # create distance matrix
  distance_matrix <- st_distance(x, x_lead)

  # diagonal of the distance matrix is your desired output, fill last entry with NA and 
  # unit
  c(diag(distance_matrix), set_units(NA, "US_survey_foot"))

}

# group by street and calculate distance
intersections <- group_by(intersections, street) %>%
  mutate(travel_distance = find_Distance(geometry))

# if needed, set unit of travel distance
units(intersections$travel_distance) <- as_units("US_survey_foot")

推荐阅读