我在R中做了一个自定义函数，使用R基本包中可用的函数来计算两个空间点之间的距离(km)。

custom_hav_dist <- function(lat1, lon1, lat2, lon2) {
R <- 6371
Radian_factor <- 0.0174533
lat_1 <- (90-lat1)*Radian_factor
lat_2 <- (90-lat2)*Radian_factor
diff_long <-(lon1-lon2)*Radian_factor

distance_in_km <- 6371*acos((cos(lat_1)*cos(lat_2))+ 
                 (sin(lat_1)*sin(lat_2)*cos(diff_long)))
rm(lat1, lon1, lat2, lon2)
return(distance_in_km)
}

样例输出

custom_hav_dist(50.31,19.08,54.14,19.39)
[1] 426.3987

PS:要计算以英里为单位的距离，请将函数R(6371)替换为3958.756(海里使用3440.065)。

由于这是关于这个话题最受欢迎的讨论，我将在这里补充我从2019年底到2020年初的经验。为了补充现有的答案-我的重点是找到一个准确和快速(即向量化)的解决方案。

让我们从这里最常用的答案——哈弗辛方法开始。向量化是很简单的，参见下面python中的例子:

def haversine(lat1, lon1, lat2, lon2):
    """
    Calculate the great circle distance between two points
    on the earth (specified in decimal degrees)

    All args must be of equal length.
    Distances are in meters.
    
    Ref:
    https://stackoverflow.com/questions/29545704/fast-haversine-approximation-python-pandas
    https://ipython.readthedocs.io/en/stable/interactive/magics.html
    """
    Radius = 6.371e6
    lon1, lat1, lon2, lat2 = map(np.radians, [lon1, lat1, lon2, lat2])

    dlon = lon2 - lon1
    dlat = lat2 - lat1

    a = np.sin(dlat/2.0)**2 + np.cos(lat1) * np.cos(lat2) * np.sin(dlon/2.0)**2

    c = 2 * np.arcsin(np.sqrt(a))
    s12 = Radius * c
    
    # initial azimuth in degrees
    y = np.sin(lon2-lon1) * np.cos(lat2)
    x = np.cos(lat1)*np.sin(lat2) - np.sin(lat1)*np.cos(lat2)*np.cos(dlon)
    azi1 = np.arctan2(y, x)*180./math.pi

    return {'s12':s12, 'azi1': azi1}

就精确度而言，它是最不准确的。维基百科在没有任何来源的情况下表示相对偏差平均为0.5%。我的实验显示偏差较小。以下是10万个随机点与我的库的比较，应该精确到毫米级:

np.random.seed(42)
lats1 = np.random.uniform(-90,90,100000)
lons1 = np.random.uniform(-180,180,100000)
lats2 = np.random.uniform(-90,90,100000)
lons2 = np.random.uniform(-180,180,100000)
r1 = inverse(lats1, lons1, lats2, lons2)
r2 = haversine(lats1, lons1, lats2, lons2)
print("Max absolute error: {:4.2f}m".format(np.max(r1['s12']-r2['s12'])))
print("Mean absolute error: {:4.2f}m".format(np.mean(r1['s12']-r2['s12'])))
print("Max relative error: {:4.2f}%".format(np.max((r2['s12']/r1['s12']-1)*100)))
print("Mean relative error: {:4.2f}%".format(np.mean((r2['s12']/r1['s12']-1)*100)))

输出:

Max absolute error: 26671.47m
Mean absolute error: -2499.84m
Max relative error: 0.55%
Mean relative error: -0.02%

因此，在10万对随机坐标上，平均偏差为2.5km，这可能对大多数情况都是好的。

下一个选择是Vincenty公式，精确到毫米，这取决于收敛标准，也可以向量化。它确实有在对跖点附近收敛的问题。你可以通过放宽收敛标准使其收敛于这些点，但准确度会下降到0.25%甚至更多。在对映点之外，Vincenty将提供与地理库相近的结果，相对误差小于1。平均是E-6。

这里提到的Geographiclib实际上是当前的黄金标准。它有几个实现，而且相当快，特别是如果你使用的是c++版本。

Now, if you are planning to use Python for anything above 10k points I'd suggest to consider my vectorized implementation. I created a geovectorslib library with vectorized Vincenty routine for my own needs, which uses Geographiclib as fallback for near antipodal points. Below is the comparison vs Geographiclib for 100k points. As you can see it provides up to 20x improvement for inverse and 100x for direct methods for 100k points and the gap will grow with number of points. Accuracy-wise it will be within 1.e-5 rtol of Georgraphiclib.

Direct method for 100,000 points
94.9 ms ± 25 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
9.79 s ± 1.4 s per loop (mean ± std. dev. of 7 runs, 1 loop each)

Inverse method for 100,000 points
1.5 s ± 504 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
24.2 s ± 3.91 s per loop (mean ± std. dev. of 7 runs, 1 loop each)

下面是VB的实现。NET，这个实现将根据您传递的Enum值以KM或Miles为单位给您结果。

Public Enum DistanceType
    Miles
    KiloMeters
End Enum

Public Structure Position
    Public Latitude As Double
    Public Longitude As Double
End Structure

Public Class Haversine

    Public Function Distance(Pos1 As Position,
                             Pos2 As Position,
                             DistType As DistanceType) As Double

        Dim R As Double = If((DistType = DistanceType.Miles), 3960, 6371)

        Dim dLat As Double = Me.toRadian(Pos2.Latitude - Pos1.Latitude)

        Dim dLon As Double = Me.toRadian(Pos2.Longitude - Pos1.Longitude)

        Dim a As Double = Math.Sin(dLat / 2) * Math.Sin(dLat / 2) + Math.Cos(Me.toRadian(Pos1.Latitude)) * Math.Cos(Me.toRadian(Pos2.Latitude)) * Math.Sin(dLon / 2) * Math.Sin(dLon / 2)

        Dim c As Double = 2 * Math.Asin(Math.Min(1, Math.Sqrt(a)))

        Dim result As Double = R * c

        Return result

    End Function

    Private Function toRadian(val As Double) As Double

        Return (Math.PI / 180) * val

    End Function

End Class

下面是postgres SQL中的一个示例(以公里为单位，为英里版本，将1.609344替换为0.8684版本)

CREATE OR REPLACE FUNCTION public.geodistance(alat float, alng float, blat  

float, blng  float)
  RETURNS float AS
$BODY$
DECLARE
    v_distance float;
BEGIN

    v_distance = asin( sqrt(
            sin(radians(blat-alat)/2)^2 
                + (
                    (sin(radians(blng-alng)/2)^2) *
                    cos(radians(alat)) *
                    cos(radians(blat))
                )
          )
        ) * cast('7926.3352' as float) * cast('1.609344' as float) ;


    RETURN v_distance;
END 
$BODY$
language plpgsql VOLATILE SECURITY DEFINER;
alter function geodistance(alat float, alng float, blat float, blng float)
owner to postgres;

我通过简化公式来简化计算。

下面是Ruby版本:

include Math
earth_radius_mi = 3959
radians = lambda { |deg| deg * PI / 180 }
coord_radians = lambda { |c| { :lat => radians[c[:lat]], :lng => radians[c[:lng]] } }

# from/to = { :lat => (latitude_in_degrees), :lng => (longitude_in_degrees) }
def haversine_distance(from, to)
  from, to = coord_radians[from], coord_radians[to]
  cosines_product = cos(to[:lat]) * cos(from[:lat]) * cos(from[:lng] - to[:lng])
  sines_product = sin(to[:lat]) * sin(from[:lat])
  return earth_radius_mi * acos(cosines_product + sines_product)
end

在其他答案中，r中的实现是缺失的。

用地质圈包中的distm函数计算两点之间的距离非常简单:

distm(p1, p2, fun = distHaversine)

地点:

p1 = longitude/latitude for point(s)
p2 = longitude/latitude for point(s)
# type of distance calculation
fun = distCosine / distHaversine / distVincentySphere / distVincentyEllipsoid 

由于地球不是完美的球形，所以椭球体的文森提公式可能是计算距离的最佳方法。因此，在地质圈包中，您可以使用:

distm(p1, p2, fun = distVincentyEllipsoid)

当然，你不一定要使用geosphere包，你也可以用一个函数来计算以R为基底的距离:

hav.dist <- function(long1, lat1, long2, lat2) {
  R <- 6371
  diff.long <- (long2 - long1)
  diff.lat <- (lat2 - lat1)
  a <- sin(diff.lat/2)^2 + cos(lat1) * cos(lat2) * sin(diff.long/2)^2
  b <- 2 * asin(pmin(1, sqrt(a))) 
  d = R * b
  return(d)
}