Dan_Patterson

Standard Distance, inter-point distance:  ... the "Special Analyst" to the rescue

Blog Post created by Dan_Patterson Champion on Jul 11, 2018

The usual.... Another distance question.  This one is a little bit different.  Either the standard distance (a statistical measure in the Spatial Statistics toolset) or a distance matrix and its parameters were needed.

Why not do both without the extension or an advanced license. 

Totally supported since python and numpy and the tools to work with them are provided for you.

Note code savvy????  You can always purchase what you need.

 

Let's start with the scenario....

You can see the points that are within the polygons.  Objective! either determine the distance matrix of the points contained within each polygon AND/OR calculate the standard distance similarly.

 

: -----------------------------------------------------------------------------------------------------------------------------------------------------

In short... 

Here is the workflow.

  1. Make your imports.  Arraytools is stuff I have written (on GitHub in sortof organized form)
  2. Select your input featureclass which is created by 'Intersect' ing your points and the polygon.  That way you can bring over your attributes into the resultant point file.
  3. Make a structured array from the featureclass just bringing over the X, Y and an identifier field.
  4. Sort the result in step 3 by the identifier field... this just makes splitting it easier.
  5. Split the array into subarrays for processing.
  6. Lines 10 -19 are just for formatting the output.
  7. Line 21 and on, is the processing steps.
    1. for each subarray, get the X,Y coordinates (line 23)
    2. calculate the center of the point distribution
    3. calculate the variance for X and Y
    4. Use the above to determine the Standard Distance (1 std level, multiply by 2 or 3 for 2 std)
    5. Determine the interpoint distance matrix (line 28) returning an array with the diagonal set to zero (point to itself) and redundant (duplicate) calculations set to zero (line 29)
    6. From the nonzero values (there should be no duplicate points anyway, BTW), calculate the distance's statistical parameters.
    7. Fancy print the results

 

import arcpy
import arraytools as art

in_fc = r"C:\Path_to_Your\File_geodatabase.gdb\pnts_intersect_polygons"

a = arcpy.da.FeatureClassToNumPyArray(in_fc, ['SHAPE@X', 'SHAPE@Y', 'ID_poly'])
a_sort = np.sort(a, order='ID_poly')
a_split = art.split_array(a_sort, fld='ID_poly')

frmt = """
Group .... {}
center ... {}
standard distance ... {}
distance matrix...{}
distance results...  mean {}, std dev. {}  min  {}, max      {}
"""


# ---- let's role ----
for i in range(len(a_split)):
    a0 = a_split[i][['SHAPE@X', 'SHAPE@Y']]
    a0_xy = art.geom._view_(a0)
    cent = np.mean(a0_xy, axis=0)
    var_x = np.var(a0_xy[:, 0])
    var_y = np.var(a0_xy[:, 1])
    stand_dist = np.sqrt(var_x + var_y)
    dm = art.geom.e_dist(a0_xy, a0_xy)
    dm_result = np.tril(dm, -1)
    dm_vals = dm_result[np.nonzero(dm_result)]
    args = [i,
            cent,
            stand_dist,
            art.form_(dm_result, deci=1, prn=False),
            dm_vals.mean(),
            dm_vals.std(),
            dm_vals.min(),
            dm_vals.max()]
    print(frmt.format(*args))

 

Here is the output for the first polygon.

Group .... 0
center ... [ 304932.447  5029991.887]
standard distance ... 795.1178530213251
distance matrix...

Array... ndim: 2  shape: (24, 24)
. .     0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   253.6    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   179.9  214.1    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   377.4  417.6  225.6    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   644.3  513.3  466.6  379.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   866.3  685.6  697.3  637.8  259.5    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .   819.0  696.3  639.5  514.0  183.1  219.6    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .  1289.9 1164.9 1110.1  960.7  654.6  527.5  473.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0.....
. .  1431.1 1290.0 1251.8 1113.8  788.6  622.9  612.4  164.4    0.0    0.0    0.0    0.0    0.0    0.0.....
. .  1583.5 1385.9 1416.1 1340.3  965.7  718.9  836.0  567.6  446.2    0.0    0.0    0.0    0.0    0.0.....
. .  1504.4 1410.2 1326.3 1148.0  897.6  810.8  715.9  292.8  309.6  745.2    0.0    0.0    0.0    0.0.....
. .  1777.6 1566.1 1616.3 1556.9 1178.8  924.2 1061.2  801.3  669.2  237.4  953.7    0.0    0.0    0.0.....
. .  1587.4 1483.2 1408.4 1236.2  969.9  861.3  786.9  333.8  297.2  703.4  101.2  898.4    0.0    0.0.....
. .  1643.8 1489.1 1465.6 1336.7  999.5  807.5  829.1  391.9  228.1  368.6  429.3  540.1  361.5    0.0.....
. .  1825.6 1740.4 1649.0 1461.7 1228.7 1138.9 1047.5  613.0  571.6  932.6  332.2 1096.0  282.0  565.3.....
. .  2069.4 1859.9 1906.4 1838.6 1462.6 1211.4 1335.0 1018.9  866.6  499.1 1108.9  294.1 1033.9  680.2.....
. .  2218.6 2007.6 2056.0 1988.4 1612.5 1361.3 1484.1 1157.9 1002.1  648.5 1229.6  441.6 1149.2  804.7.....
. .  2004.9 1880.9 1825.1 1663.3 1371.9 1220.5 1190.2  717.4  597.6  791.3  533.6  889.5  435.4  454.5.....
. .  2202.6 2002.9 2034.8 1950.7 1580.4 1337.6 1439.5 1076.4  915.1  619.2 1116.6  455.6 1030.7  702.8.....
. .  2075.5 1916.9 1897.4 1766.0 1431.2 1232.4 1260.5  810.0  652.3  633.7  736.6  650.5  637.8  431.7.....
. .  2144.5 1974.1 1968.1 1848.3 1501.8 1288.7 1337.3  904.1  741.6  636.9  859.7  602.4  763.0  514.0.....
. .  2161.4 2055.3 1982.6 1806.5 1542.4 1413.2 1359.3  893.4  795.7 1021.5  658.6 1115.3  574.4  678.7.....
. .  2245.4 2089.5 2066.9 1931.0 1601.2 1405.5 1428.7  971.7  817.8  798.8  868.6  788.6  767.6  602.0.....
. .  2390.9 2267.3 2211.2 2046.4 1758.7 1603.1 1577.0 1104.2  981.0 1103.4  905.0 1140.0  811.6  813.9.....

I will put this in the Point Tools toolset at some point.  

If people need the code, email me at the link on my profile and I can direct you to the links on GitHub.

Outcomes