Kriging插值计算
参考论文: http://people.ku.edu/~gbohling/cpe940
# -*- coding: utf-8 -*-
# ---------------------------------------------------------------------------
# Kriging.py
# Created on: 2014-06-12 10:14:21.00000
# (generated by ArcGIS/ModelBuilder)
# Description:
# --------------------------------------------------------------------------- # Import arcpy module
import os
import math
import sys from pylab import *
import numpy as np
from pandas import DataFrame, Series
from scipy.spatial.distance import pdist, squareform # 计算距离 # dataXYV = [{"x":12100,"y":8300,"v":14.6515},{"x":5300,"y":8700,"v":14.5093},{"x":3500,"y":13900,"v":14.0639},{"x":5100,"y":1900,"v":15.1084},{"x":9900,"y":13700,"v":13.919},{"x":2900,"y":900,"v":13.1304},{"x":7900,"y":6700,"v":14.5724},{"x":16900,"y":4900,"v":15.0814},{"x":18700,"y":1500,"v":13.91},{"x":2700,"y":2100,"v":13.4024},{"x":10700,"y":5100,"v":14.9395},{"x":7500,"y":12900,"v":15.2159},{"x":5500,"y":11100,"v":14.5777},{"x":9500,"y":9100,"v":14.2483},{"x":15300,"y":3100,"v":14.4281},{"x":4700,"y":9700,"v":15.2606},{"x":16700,"y":15700,"v":16.1859},{"x":19500,"y":9700,"v":14.2079},{"x":16900,"y":13100,"v":16.9583},{"x":900,"y":3700,"v":13.8354},{"x":500,"y":11900,"v":14.1859},{"x":9100,"y":1300,"v":14.0381},{"x":9100,"y":13700,"v":14.3685},{"x":9900,"y":12900,"v":13.4018},{"x":6300,"y":100,"v":15.8953},{"x":3700,"y":5100,"v":12.8667},{"x":16300,"y":900,"v":15.1039},{"x":18300,"y":13500,"v":15.7736},{"x":9500,"y":6900,"v":14.1333},{"x":17900,"y":3100,"v":13.3369},{"x":9900,"y":15500,"v":15.1362},{"x":7100,"y":8900,"v":15.0847},{"x":19300,"y":7100,"v":14.2498},{"x":2300,"y":5700,"v":12.6811},{"x":7300,"y":8900,"v":14.9384},{"x":13900,"y":3700,"v":15.6005},{"x":8500,"y":10100,"v":13.7796},{"x":8100,"y":8700,"v":15.2907},{"x":14700,"y":11900,"v":15.6881},{"x":6300,"y":2300,"v":15.3677},{"x":11900,"y":12900,"v":14.3283},{"x":18100,"y":7100,"v":14.7374},{"x":11300,"y":7100,"v":15.0547},{"x":12500,"y":3100,"v":14.8889},{"x":2700,"y":12700,"v":14.436},{"x":2700,"y":4300,"v":12.1491},{"x":8500,"y":11300,"v":13.624},{"x":1500,"y":900,"v":14.188},{"x":7300,"y":1300,"v":14.9072},{"x":10700,"y":4100,"v":15.2029},{"x":7100,"y":1900,"v":15.3468},{"x":3900,"y":8500,"v":15.939},{"x":17100,"y":6100,"v":15.7269},{"x":14100,"y":10100,"v":15.3238},{"x":11500,"y":4900,"v":14.0445},{"x":13300,"y":15700,"v":14.4032},{"x":1900,"y":12100,"v":14.3586},{"x":15100,"y":2900,"v":14.6007},{"x":6500,"y":900,"v":16.1458},{"x":8900,"y":6100,"v":15.7727},{"x":4500,"y":2300,"v":13.6234},{"x":12900,"y":10300,"v":15.1024},{"x":10900,"y":5700,"v":15.3546},{"x":3500,"y":700,"v":13.8431},{"x":16300,"y":3700,"v":14.9427},{"x":900,"y":5100,"v":14.4139},{"x":12900,"y":12900,"v":13.6177},{"x":15300,"y":9300,"v":16.3787},{"x":7300,"y":6900,"v":14.258},{"x":16300,"y":12500,"v":15.7772},{"x":100,"y":8900,"v":14.6553},{"x":1700,"y":11700,"v":14.3627},{"x":17500,"y":11100,"v":15.9659},{"x":14900,"y":8300,"v":16.0095},{"x":8300,"y":10900,"v":13.9639},{"x":4100,"y":14500,"v":14.2649},{"x":11100,"y":15300,"v":15.7684},{"x":500,"y":4900,"v":14.591},{"x":13100,"y":1500,"v":15.1377},{"x":18900,"y":1700,"v":14.095},{"x":3500,"y":7500,"v":15.1486},{"x":3700,"y":6900,"v":13.9584},{"x":14500,"y":13300,"v":14.7381},{"x":4900,"y":9100,"v":15.0689},{"x":9700,"y":5700,"v":15.8042}]
dataXYV = [{"x":12100.00,"y":8300.00,"v":14.6515},
{"x":5300.00,"y":8700.00,"v":14.5093},
{"x":3500.00,"y":13900.00,"v":14.0639},
{"x":5100.00,"y":1900.00,"v":15.1084},
{"x":9900.00,"y":13700.00,"v":13.919},
{"x":2900.00,"y":900.00,"v":13.1304},
{"x":7900.00,"y":6700.00,"v":14.5724},
{"x":16900.00,"y":4900.00,"v":15.0814},
{"x":18700.00,"y":1500.00,"v":13.91},
{"x":2700.00,"y":2100.00,"v":13.4024},
{"x":10700.00,"y":5100.00,"v":14.9395},
{"x":7500.00,"y":12900.00,"v":15.2159},
{"x":5500.00,"y":11100.00,"v":14.5777},
{"x":9500.00,"y":9100.00,"v":14.2483},
{"x":15300.00,"y":3100.00,"v":14.4281},
{"x":4700.00,"y":9700.00,"v":15.2606},
{"x":16700.00,"y":15700.00,"v":16.1859},
{"x":19500.00,"y":9700.00,"v":14.2079},
{"x":16900.00,"y":13100.00,"v":16.9583},
{"x":900.00,"y":3700.00,"v":13.8354},
{"x":500.00,"y":11900.00,"v":14.1859},
{"x":9100.00,"y":1300.00,"v":14.0381},
{"x":9100.00,"y":13700.00,"v":14.3685},
{"x":9900.00,"y":12900.00,"v":13.4018},
{"x":6300.00,"y":100.00,"v":15.8953},
{"x":3700.00,"y":5100.00,"v":12.8667},
{"x":16300.00,"y":900.00,"v":15.1039},
{"x":18300.00,"y":13500.00,"v":15.7736},
{"x":9500.00,"y":6900.00,"v":14.1333},
{"x":17900.00,"y":3100.00,"v":13.3369},
{"x":9900.00,"y":15500.00,"v":15.1362},
{"x":7100.00,"y":8900.00,"v":15.0847},
{"x":19300.00,"y":7100.00,"v":14.2498},
{"x":2300.00,"y":5700.00,"v":12.6811},
{"x":7300.00,"y":8900.00,"v":14.9384},
{"x":13900.00,"y":3700.00,"v":15.6005},
{"x":8500.00,"y":10100.00,"v":13.7796},
{"x":8100.00,"y":8700.00,"v":15.2907},
{"x":14700.00,"y":11900.00,"v":15.6881},
{"x":6300.00,"y":2300.00,"v":15.3677},
{"x":11900.00,"y":12900.00,"v":14.3283},
{"x":18100.00,"y":7100.00,"v":14.7374},
{"x":11300.00,"y":7100.00,"v":15.0547},
{"x":12500.00,"y":3100.00,"v":14.8889},
{"x":2700.00,"y":12700.00,"v":14.436},
{"x":2700.00,"y":4300.00,"v":12.1491},
{"x":8500.00,"y":11300.00,"v":13.624},
{"x":1500.00,"y":900.00,"v":14.188},
{"x":7300.00,"y":1300.00,"v":14.9072},
{"x":10700.00,"y":4100.00,"v":15.2029},
{"x":7100.00,"y":1900.00,"v":15.3468},
{"x":3900.00,"y":8500.00,"v":15.939},
{"x":17100.00,"y":6100.00,"v":15.7269},
{"x":14100.00,"y":10100.00,"v":15.3238},
{"x":11500.00,"y":4900.00,"v":14.0445},
{"x":13300.00,"y":15700.00,"v":14.4032},
{"x":1900.00,"y":12100.00,"v":14.3586},
{"x":15100.00,"y":2900.00,"v":14.6007},
{"x":6500.00,"y":900.00,"v":16.1458},
{"x":8900.00,"y":6100.00,"v":15.7727},
{"x":4500.00,"y":2300.00,"v":13.6234},
{"x":12900.00,"y":10300.00,"v":15.1024},
{"x":10900.00,"y":5700.00,"v":15.3546},
{"x":3500.00,"y":700.00,"v":13.8431},
{"x":16300.00,"y":3700.00,"v":14.9427},
{"x":900.00,"y":5100.00,"v":14.4139},
{"x":12900.00,"y":12900.00,"v":13.6177},
{"x":15300.00,"y":9300.00,"v":16.3787},
{"x":7300.00,"y":6900.00,"v":14.258},
{"x":16300.00,"y":12500.00,"v":15.7772},
{"x":100.00,"y":8900.00,"v":14.6553},
{"x":1700.00,"y":11700.00,"v":14.3627},
{"x":17500.00,"y":11100.00,"v":15.9659},
{"x":14900.00,"y":8300.00,"v":16.0095},
{"x":8300.00,"y":10900.00,"v":13.9639},
{"x":4100.00,"y":14500.00,"v":14.2649},
{"x":11100.00,"y":15300.00,"v":15.7684},
{"x":500.00,"y":4900.00,"v":14.591},
{"x":13100.00,"y":1500.00,"v":15.1377},
{"x":18900.00,"y":1700.00,"v":14.095},
{"x":3500.00,"y":7500.00,"v":15.1486},
{"x":3700.00,"y":6900.00,"v":13.9584},
{"x":14500.00,"y":13300.00,"v":14.7381},
{"x":4900.00,"y":9100.00,"v":15.0689},
{"x":9700.00,"y":5700.00,"v":15.8042}] length = len(dataXYV)
distanceMatrix = [[] for i in range(length)]
index = 0
distTotal = 0;
distMin = 1.0e15
distMax = -1.0e15
distAver = 0
for x in dataXYV:
for y in dataXYV:
z = math.sqrt((x['x']-y['x'])*(x['x']-y['x'])+(x['y']-y['y'])*(x['y']-y['y']))
distTotal += z
if z > distMax:
distMax = z
if z < distMin and x != y:
distMin = z
distanceMatrix[index].append(z)
index += 1
distAver = distTotal / (length * length - length)
dataInfo = {'count':(length * length - length), 'distAver':distAver, 'distMin':distMin, 'distMax':distMax}
#print dataInfo '''
for i in range(0, length):
for j in range(0, length):
print(int(lists[i][j])),
print(';')
''' '''
查找点对
'''
def findPairs(dataXYV, distanceMatrix, minValue, maxValue):
totalDistance = 0;
count = 0;
minDistance = 1.0e15
maxDistance = -1.0e15
averageDistance = 0
pairs = []
for i in range(0, length):
for j in range(i+1, length):
if distanceMatrix[i][j]>minValue and distanceMatrix[i][j]<=maxValue:
#if math.fabs(dataXYV[i]['x']-dataXYV[j]['x'])>minValue and math.fabs(dataXYV[i]['y']-dataXYV[j]['y'])>minValue and math.fabs(dataXYV[i]['x']-dataXYV[j]['x'])<=maxValue and math.fabs(dataXYV[i]['y']-dataXYV[j]['y'])<=maxValue:
# print(int(lists[i][j])),
totalDistance += distanceMatrix[i][j]
count += 1
if distanceMatrix[i][j] >= maxDistance:
maxDistance = distanceMatrix[i][j]
if distanceMatrix[i][j] <= minDistance:
minDistance = distanceMatrix[i][j]
pair = {'i':i,'j':j,'iv':dataXYV[i]['v'],'jv':dataXYV[j]['v'],'dist':distanceMatrix[i][j]}
pairs.append(pair)
#print(count)
averageDistance = totalDistance / count
info = {'count':count, 'distAver':averageDistance, 'distMin':minDistance, 'distMax':maxDistance}
#print info
#print pairs
return pairs '''
计算统计信息: 协方差、相关系数、半方变异
'''
def computeStatisticInfo(pairs):
pairCount = len(pairs)
distanceTotal = 0
distanceAverage = 0
#
v1v2Total=0
v1Total=0
v2Total=0
#
v1v1Total=0
v2v2Total=0
#
v1_v2Total=0
#
for x in pairs:
val1 = x['iv']
val2 = x['jv']
distanceTotal = distanceTotal + x['dist']
v1v2Total = v1v2Total + val1 * val2
v1Total = v1Total + val1
v2Total = v2Total + val2
#
v1v1Total = v1v1Total + val1 * val1
v2v2Total = v2v2Total + val2 * val2
#
v1_v2Total = v1_v2Total + math.pow(val1 - val2, 2)
#
distanceAverage = distanceTotal / pairCount
v1v2Covariance = v1v2Total / pairCount - v1Total * v2Total / (pairCount * pairCount)
v1v2Corelation = (v1v2Total*pairCount - v1Total * v2Total) / math.sqrt(v1v1Total * pairCount - v1Total * v1Total) / math.sqrt(v2v2Total * pairCount - v2Total * v2Total)
v1v2Semivariance = v1_v2Total / (pairCount * 2)
statisticInfo = {'covariance':v1v2Covariance, 'corelation':v1v2Corelation, 'semivariance':v1v2Semivariance, 'count':pairCount, 'distAver':distanceAverage} #
# print statisticInfo
return statisticInfo '''
计算各种lagSize下的统计信息: 协方差、相关系数、半方变异
'''
def staticInfoAll(dataXYV, distanceMatrix, lagCellSize, lagCount):
semiLagCellSize = lagCellSize / 2
pairsStaticInfos = []
for i in range(0, lagCount-1): lagSize = lagCellSize * i
lagMin = lagSize - semiLagCellSize
if lagMin < 0:
lagMin = 0
lagMax = lagSize + semiLagCellSize
#print(lagMin, lagMax, lagSize) '''
lagMin = lagCellSize * i
lagMax = lagCellSize * (i + 1)
lagSize = (lagCellSize * i + lagCellSize * (i + 1))/2
#print lagMin, lagMax, lagSize
'''
pairs = findPairs(dataXYV, distanceMatrix, lagMin, lagMax)
statisticInfo = computeStatisticInfo(pairs)
statisticInfo['lagSize'] = lagSize
print(lagMin, lagMax, statisticInfo['lagSize'], statisticInfo['count'], statisticInfo['distAver'], statisticInfo['covariance'], statisticInfo['corelation'], statisticInfo['semivariance'])
pairsStaticInfos.append(statisticInfo)
return pairsStaticInfos
'''
def computeC0(dataXYV):
valueTotal = 0;
valueAver = 0;
variance = 0;
for x in dataXYV:
valueTotal += x['v']
valueAver = valueTotal / length
print valueAver
for x in dataXYV:
variance += math.pow((x['v'] - valueAver),2)
variance = variance / length
# variance = math.sqrt(variance)
print variance computeC0(dataXYV)
''' def optimization(pairsStaticInfos):
aMin = 0
cMin = 0
varianceTotalMin = 1.0e45
hSize = len(pairsStaticInfos)
for aValue in range( 3500, 4500):
for c1Value in range(60, 99):
cValue = c1Value / 100.0
#cValue = 0.78
#print(aValue, cValue)
#print(aValue, cValue)
varianceTotal = 0
for x in pairsStaticInfos:
y = spherical( x['distAver'], aValue, cValue ) # distAver lagSize
#print y, x['semivariance']
varianceTotal = varianceTotal + ((y - x['semivariance'])**2.0) * x['count']
varianceTotal = varianceTotal / hSize
#print varianceTotal
if varianceTotal <= varianceTotalMin:
varianceTotalMin = varianceTotal
aMin = aValue
cMin = cValue
#print(aMin, cMin, varianceTotalMin) para = {"a":aMin, "c0":cMin}
print(para)
return para def spherical( h, a, C0):
if h <= a:
return (C0*( 1.5*h/a - 0.5*(h/a)**3.0 ))
else:
return C0 pairsStaticInfos = staticInfoAll(dataXYV, distanceMatrix, 1000, 12) # 500, 25 para = optimization(pairsStaticInfos) lagSize = []
semivariance = []
modelY = []
modelYY = []
#
lagSize.append(0)
semivariance.append(0)
modelY.append(0)
for x in pairsStaticInfos:
y = spherical( x['distAver'], para['a'], para['c0'] )
#yy= spherical( x['distAver'], 4141, para['c0'] )
lagSize.append(x['distAver'])
semivariance.append(x['semivariance'])
modelY.append(y)
modelYY.append(yy) plot( lagSize, semivariance, 'o' )
plot( lagSize, modelY, '.-' ) ;
title('Spherical Model')
ylabel('Semivariance')
xlabel('Lag [m]')
Kriging插值计算的更多相关文章
- Matlab插值计算各时刻磁法勘探日变观测值
Matlab插值计算各时刻磁法勘探日变观测值 在磁法勘探中,消日变影响的改正称为日变改正.进行日变改正时必须设立日变站,观测日变情况.根据日变数据和测点观测时间,对观测数据进行改正. 在本次磁法实习中 ...
- (转)Unity3D 之插值计算
在unity3D中经常用线性插值函数Lerp()来在两者之间插值,两者之间可以是两个材质之间.两个向量之间.两个浮点数之间.两个颜色之间,其函数原型如下: Material.Lerp 插值 funct ...
- Swift字符串插值
字符串插值是一种全新的构建字符串的方式,可以在其中包含常量.变量.字面量和表达式.您插入的字符串字面量的每一项都被包裹在以反斜线为前缀的圆括号中: let multiplier = let messa ...
- Tetrahedron based light probe interpolation(基于四面体的Light Probe插值)
在当前的游戏引擎中,使用Light Probe来计算全局环境光对于动态物体的影响是一种很主流的方法.在预处理阶段生成完场景的Light Probe之后,传统的方法采用查找最近的8个相邻的Probe然后 ...
- 插值技术之Bezier插值(1) -- Bezier Curve
作者:i_dovelemon 来源:CSDN 日期:2015 / 7 / 11 主题:Interpolate,Bezier Curve 引言 在游戏开发中.诸如动画系统.路径计算等等操作,都会遇到对数 ...
- 基于GDAL的栅格图像空间插值预处理
转自 基于GDAL的栅格图像空间插值预处理——C语言版 基于GDAL的栅格图像预处理 前言 栅格数据和矢量数据构成空间数据的主要来源,怎样以开源方式读取并处理这些空间数据?目前有多种开源支持包,这里只 ...
- 从零开始一起学习SLAM | 用四元数插值来对齐IMU和图像帧
视觉 Vs. IMU 小白:师兄,好久没见到你了啊,我最近在看IMU(Inertial Measurement Unit,惯性导航单元)相关的东西,正好有问题求助啊 师兄:又遇到啥问题啦? 小白:是这 ...
- three.js使用gpu选取物体并计算交点位置
光线投射法 使用three.js自带的光线投射器(Raycaster)选取物体非常简单,代码如下所示: var raycaster = new THREE.Raycaster(); var mouse ...
- 【bzoj4559】[JLoi2016]成绩比较（dp+拉格朗日插值）
bzoj 题意: 有\(n\)位同学,\(m\)门课. 一位同学在第\(i\)门课上面获得的分数上限为\(u_i\). 定义同学\(A\)碾压同学\(B\)为每一课\(A\)同学的成绩都不低于\(B\ ...
随机推荐
- 【xsy1504】 pitcure 树状数组
数据范围:$n≤2\times 10^5$ 以下是题解: #include<bits/stdc++.h> #define L unsigned int #define MOD 167772 ...
- 【xsy1156】 树套树(tree) 倍增
题目大意:给你$m$棵由$n$个点构成的全等的树$A$.这$m$棵树之间有$m-1$条边相连,组成了一棵大树. 有$q$组询问,每次询问这棵大树上两点之间的距离. $n,m,q≤10^5$ 这是一道小 ...
- POJ 1056
#include <iostream> #include <string> #define MAXN 50 using namespace std; struct node { ...
- Android UiAutomator UiDevice API
UiDevice为单例模式 1.获取设备 static UiDevice getInstance() This method is deprecated. Should use getInstance ...
- c# Xml反序列化示例
Xml 示例 <?xml version="1.0" encoding="utf-8"?> <CarCollection> <Ca ...
- Zookeeper在Centos7上搭建单节点应用
(默认机器上已经安装并配置好了jdk) 1.下载zookeeper并解压 $ tar -zxvf zookeeper-3.4.6.tar.gz 2.将解压后的文件夹移动到 /usr/local/ 目录 ...
- android学习-LocationManager（一）-
Location Provider是不同位置信息来源的抽象 Location封装了从位置提供者提供给应用的位置数据(经纬度) Criteria提供了查询获取包含特定特征的位置提供者(accuracy精 ...
- SPP(Spatial Pyramid Pooling)详解
一直对Fast RCNN中ROI Pooling层不解,不同大小的窗口输入怎么样才能得到同样大小的窗口输出呢,今天看到一篇博文讲得挺好的,摘录一下,方便查找. Introduction 在一般的CNN ...
- （转）linux内核调优参数对比和解释
[net] ######################## cat /proc/sys/net/ipv4/tcp_syncookies # 默认值:1 # 作用:是否打开SYN Cookie功能,该 ...
- 机器学习入门学习笔记：（一）BP神经网络原理推导及程序实现
机器学习中,神经网络算法可以说是当下使用的最广泛的算法.神经网络的结构模仿自生物神经网络,生物神经网络中的每个神经元与其他神经元相连,当它“兴奋”时,想下一级相连的神经元发送化学物质,改变这些神经元的 ...