{smcl}
{* *! version 1.0.6  16may2012 author: Thomas Grund}{...}
{marker topic}
{helpb nw_topical##analysis:[NW-2.6] Analysis}

{title:Title}

{p2colset 9 19 22 2}{...}
{p2col :nwcontext {hline 2} Create a context variable}
{p2colreset}{...}

{title:Syntax}

{p 8 17 2}
{cmdab: nwcontext} 
[{it:{help netname}}]
{cmd:,}
{opth attribute(varname)}
[{opt stat}({it:{help nwcontext##statistic:statistic}})
{opt mode}({it:{help nwcontext##context:context}})
{opth generate(newvarname)}
{opth mat(string)}
{opt noweight}]



{synoptset 25 tabbed}{...}
{synopthdr}
{synoptline}
{synopt:{opth attribute(varname)}}attribute variable{p_end}
{synopt:{opt stat}({it:{help nwcontext##statistic:statistic}})}statistic that is used to calculate context variable for node i from attributes of network neighbors{p_end}
{synopt:{opt mode}({it:{help nwcontext##context:context}})}defines network neighbors of node i as either nodes j who receive ties from i, send ties to j or both{p_end}
{synopt:{opth generate(newvarname)}}name of the context variable to be generated; default = {it:_context_varname}{p_end}
{synopt:{opth mat(string)}}name of new mata matrix where context variable should be stored instead{p_end}
{synopt:{opt noweight}}ignores valued ties and treats all as binary{p_end}

{synoptline}
{p2colreset}{...}

{synoptset 20 tabbed}{...}
{marker statistic}{...}
{p2col:{it:statistic}}Description{p_end}
{p2line}
{p2col:{cmd: mean}}mean of {help varname} over network neighbors; default
		{p_end}
{p2col:{cmd: max}}maximum of {help varname} over network neighbors
		{p_end}
{p2col:{cmd: min}}minimum of {help varname} over network neighbors
		{p_end}
{p2col:{cmd: sum}}sum of {help varname} over network neighbors
		{p_end}
{p2col:{cmd: sd}}standard deviation of {help varname} over network neighbors
		{p_end}
{p2col:{cmd: meanego}}mean of {help varname} over network neighbors and ego)
		{p_end}
{p2col:{cmd: maxego}}maximum of {help varname} over network neighbors and ego
		{p_end}
{p2col:{cmd: minego}}minimum of {help varname} over network neighbors and ego
		{p_end}
{p2col:{cmd: sumego}}sum of {help varname} over network neighbors and ego
		{p_end}
{p2col:{cmd: sdego}}standard deviation of {help varname} over network neighbors and ego
		{p_end}
		
{p2colreset}{...}
{synoptset 20 tabbed}{...}
{marker context}{...}
{p2col:{it:context}}Description{p_end}
{p2line}
{p2col:{cmd: outgoing}}network neighbors of node i are all nodes j who receive a tie from i; default
		{p_end}
{p2col:{cmd: incoming}}network neighbors of node i are all nodes j who send a tie to i; default
		{p_end}
{p2col:{cmd: both}}network neighbors of node i are all nodes j who either send a tie to i or receive a tie from i
		{p_end}


{title:Description}

{pstd}
{cmd:nwcontext} generates a context variable, which holds information about the {it:attribute values} of a node's network neighbors.

{pstd}
For each node the set of network neighbors is specified in {opt mode()}, by default {it:context} = {bf:outgoing}. This means that the new context
variable {it:newvarname}[i] is calculated based on an all values {it:varname}[j], for which there is a tie from node {it:i} to node
{it:j}. A network neighborhood of
node {it:i} can be either 1) all nodes {it:j} to whom i has outgoing ties (default), 2) all nodes {it:j} from whom {it:i} receives incoming ties, or 3) both. In the case 
of valued ties {it:varname}[j] is weighted accordingly.

{pstd}
After that a {it:statistic} defined in {opt stat()} is calculated from all weigthed {it:varname}[j]. By default, the mean is calculated. More formally:

{pmore}
{it:newvarname}[i] = {it:stat}({it:varname}[j]), for all {it:j} with {it:y_ij} > 0

{pstd}
When used with the defaults, the command calculates for each node {it:i} the mean score of its network neighbors on {help varname}.
Valid statistics are: the{bf: mean}, the {bf:max}, the {bf:min}, the {bf:sum} and the {bf:sd} of {help varname}.

{pstd}
Sometimes, however, one might want to calculate statistics including the attribute of ego. One can achieve this by using the ego-extended version of each statistic
({bf:meanego},  {bf:maxego},  {bf:minego},  {bf:sumego} and {bf:sdego}), which are calculated in this way:

{pmore}
{it:newvarname}[i] = {it:stat}({it:varname}[j]), for all {it:j} with {it:y_ij} > 0 or j == i


{title:Remarks}

{pstd}
In the case of undirected networks, no {it: mode} option needs to be specified. 


{title:Examples}

{pstd}
This example loads the Florentine marriage data. The variable {it:wealth} indicates how rich each family is. {bf: nwcontext}
generates different variables: {it:w_avg} = average wealth of network neighbors,  {it:w_min} = wealth of poorest network neighbor,
{it:w_min} = wealth of richest network neighbor, {it:w_sd} = standard deviation of wealth over network neighbors.

	{com}. webnwuse florentine
	{com}. nwcontext flomarriage, attribute(wealth) generate(w_avg)
	{com}. nwcontext flomarriage, attribute(wealth) generate(w_min) stat(min)
	{com}. nwcontext flomarriage, attribute(wealth) generate(w_max) stat(max)
	{com}. nwcontext flomarriage, attribute(wealth) generate(w_sd) stat(sd)

	{com}. list w*
{txt}
     {c TLC}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c TRC}
     {c |} {res}wealth      w_avg      w_min      w_max       w_sd {txt}{c |}
     {c LT}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c RT}
  1. {c |} {res}    10        103        103        103          0 {txt}{c |}
  2. {c |} {res}    36   47.66667          8        103   40.33471 {txt}{c |}
  3. {c |} {res}    55       61.5         20        103       41.5 {txt}{c |}
  4. {c |} {res}    44   67.66666          8        146   57.86382 {txt}{c |}
  5. {c |} {res}    20   83.33334         49        146   44.37968 {txt}{c |}
     {c LT}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c RT}
  6. {c |} {res}    32         36         36         36          0 {txt}{c |}
  7. {c |} {res}     8       42.5         36         48   4.330127 {txt}{c |}
  8. {c |} {res}    42          8          8          8          0 {txt}{c |}
  9. {c |} {res}   103         31         10         55   17.26268 {txt}{c |}
 10. {c |} {res}    48         10         10         10          0 {txt}{c |}
     {c LT}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c RT}
 11. {c |} {res}    49         70         20        146     54.626 {txt}{c |}
 12. {c |} {res}     3          .          .          .          . {txt}{c |}
 13. {c |} {res}    27         99         48        146   40.10819 {txt}{c |}
 14. {c |} {res}    10       75.5         48        103       27.5 {txt}{c |}
 15. {c |} {res}   146         35         20         49   11.89538 {txt}{c |}
     {c LT}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c RT}
 16. {c |} {res}    48         46          8        103   41.04469 {txt}{c |}
     {c BLC}{hline 8}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c -}{hline 10}{c BRC}
	
{pstd}
One can plot the marriage network with {it:wealth} as node label to better understand how these values come about:

	{cmd:. nwplot flomarriage, label(wealth)}

{title:Also see}

   {help nwneighbor}