生成矩阵_如何使用Stata生成空间权重矩阵、计算Moran's I及画莫兰散点图

一、生成空间权重矩阵

1.Create a row-standardized binary spatial weights matrix assuming spherical coordinates and a distance cut-off of 10 miles(生成行标准化的二元空间权重矩阵)

. use 31_provinces_longitude&latitude,clear

.spwmatrix gecon latitude longitude, wn(wbin) wtype(bin) db(0 550) r(3958.761) row

.clear

.svmat wbin

.save wbin

2.Generate an inverse distance squared spatial weights matrix using projected latitudes and longitudes(生成距离倒数空间权重矩阵)

. use 31_provinces_longitude&latitude,clear

. spwmatrix gecon latitude longitude, wname(winvsq) wtype(inv) alpha(1) dband(0 100) cart

. clear

. svmat winvsq

. save winvsq

二、计算Moran's I

(一)计算全域Moran's I

1.使用距离倒数空间权重矩阵

. use winv.dta,clear

. spatwmat using winv.dta,name(winv) standardize

. use 31_provinces_core_variable_chn.dta,clear

. spatgsa pcrgdp_2002 pcrgdp_2003 pcrgdp_2004 pcrgdp_2005 pcrgdp_2006 pcrgdp_2007 pcrgdp_2008 pcrgdp_2009 pcrgdp_2010 pcrgdp_2011 pcrgdp_2012 pcrgdp_2013 pcrgdp_2014 pcrgdp_2015 pcrgdp_2016 ,weights(winv) moran(见图1)

. spatgsa bcs_2002 bcs_2003 bcs_2004 bcs_2005 bcs_2006 bcs_2007 bcs_2008 bcs_2009 bcs_2010 bcs_2011 bcs_2012 bcs_2013 bcs_2014 bcs_2015 bcs_2016 ,weights(winv) moran

. spatgsa ti_2002 ti_2003 ti_2004 ti_2005 ti_2006 ti_2007 ti_2008 ti_2009 ti_2010 ti_2011 ti_2012 ti_2013 ti_2014 ti_2015 ti_2016 ,weights(winv) moran

. spatgsa ts_2002 ts_2003 ts_2004 ts_2005 ts_2006 ts_2007 ts_2008 ts_2009 ts_2010 ts_2011 ts_2012 ts_2013 ts_2014 ts_2015 ts_2016 ,weights(winv) moran

图1

2.使用二元空间权重矩阵

. use wbin.dta,clear

. spatwmat using wbin.dta,name(wbin) standardize

. use 31_provinces_core_variable_chn.dta,clear

. spatgsa pcrgdp_2002 pcrgdp_2003 pcrgdp_2004 pcrgdp_2005 pcrgdp_2006 pcrgdp_2007 pcrgdp_2008 pcrgdp_2009 pcrgdp_2010 pcrgdp_2011 pcrgdp_2012 pcrgdp_2013 pcrgdp_2014 pcrgdp_2015 pcrgdp_2016 ,weights(wbin) moran(见图2)

. spatgsa bcs_2002 bcs_2003 bcs_2004 bcs_2005 bcs_2006 bcs_2007 bcs_2008 bcs_2009 bcs_2010 bcs_2011 bcs_2012 bcs_2013 bcs_2014 bcs_2015 bcs_2016 ,weights(wbin) moran(计算效果不好)

. spatgsa ti_2002 ti_2003 ti_2004 ti_2005 ti_2006 ti_2007 ti_2008 ti_2009 ti_2010 ti_2011 ti_2012 ti_2013 ti_2014 ti_2015 ti_2016 ,weights(wbin) moran

. spatgsa ts_2002 ts_2003 ts_2004 ts_2005 ts_2006 ts_2007 ts_2008 ts_2009 ts_2010 ts_2011 ts_2012 ts_2013 ts_2014 ts_2015 ts_2016 ,weights(wbin) moran(计算效果不好)

图2

(二)计算局域Moran's I

. use winv.dta,clear

. spatwmat using wbin.dta,name(winv) standardize

. use 31_provinces_core_variable_chn.dta,clear

. spatlsa pcrgdp_2002, w(winv) moran twotail(见图3)

图3

三、画Moran’s I散点图

(一)二元空间权重矩阵(0-1空间权重矩阵)

. use wbin.dta,clear

. spatwmat using wbin.dta,name(wbin) standardize

. use 31_provinces_core_variable_chn.dta,clear

. spatlsa pcrgdp_2002,weights(wbin) moran graph(moran) symbol(id) id(prvn)(见图4)

. spatlsa pcrgdp_2002,weights(wbin) moran graph(moran) symbol(id) id(prvn) twotail

图4

(二)距离倒数矩阵

. use winv.dta,clear

. spatwmat using winv.dta,name(winv) standardize

. use 31_provinces_core_variable_chn.dta,clear

. spatlsa pcrgdp_2002,weights(winv) moran graph(moran) symbol(id) id(prvn)(见图5)

. spatlsa pcrgdp_2002,weights(winv) moran graph(moran) symbol(id) id(prvn) twotail

图5

THE END!

           欢迎分享！欢迎转发！

励志与成功在统计学意义上存在显著的正相关关系，是引致成功空间溢出效应的关键因素之一！

THE WORLD IN YOUR HEART, AND TIE THE HEART TO THE BIG DIPPER！