twoway_kdensity.hlp

是一个经济学管理应用软件 很难找的 但是经济学学生又必须用到

{smcl}
{* 08apr2005}{...}
{cmd:help twoway kdensity} {right:dialogs:  {dialog twoway_overlay:overlaid twoway}{space 0}}
	{right:{dialog twoway_simple:single twoway}{space 2}}
{hline}

{title:Title}

{p2colset 5 34 36 2}{...}
{p2col :{hi:[G] graph twoway kdensity} {hline 2}}Kernel density plots{p_end}
{p2colreset}{...}


{title:Syntax}

{p 8 32 2}
{cmdab:tw:oway}
{cmd:kdensity}
	{it:varname}
	{ifin}
	{weight}
	[{cmd:,}
	{it:options}]

{p2colset 9 30 32 2}{...}
	{it:options}{col 30}description
{p2line}
	{cmdab:w:idth:(}{it:#}{cmd:)}{...}
{col 30}smoothing parameter
	{cmdab:ep:an}{...}
{col 30}use Epanechnikov kernel; the default
	{cmdab:epan2}{...}
{col 30}use alternative Epanechnikov kernel 
	{cmdab:bi:weight}{...}
{col 30}use biweight kernel
	{cmdab:cos:ine}{...}
{col 30}use cosine kernel
	{cmdab:gau:ss}{...}
{col 30}use Gaussian kernel
	{cmdab:par:zen}{...}
{col 30}use Parzen kernel
	{cmdab:rec:tangle}{...}
{col 30}use rectangular kernel
	{cmdab:tri:angle}{...}
{col 30}use triangular kernel

	{cmdab:ra:nge:(}{it:#} {it:#}{cmd:)}{...}
{col 30}range for plot, minimum and maximum
	{opth ra:nge(varname)}{...}
{col 30}range for plot obtained from {it:varname}
	{cmd:n(}{it:#}{cmd:)}{...}
{col 30}number of points to evaluate
	{cmd:area(}{it:#}{cmd:)}{...}
{col 30}rescaling parameter
	{cmdab:hor:izontal}{...}
{col 30}graph horizontally

INCLUDE help gr_clopt

INCLUDE help gr_axlnk

INCLUDE help gr_twopt
{p2line}

{phang}
{cmd:fweight}s and {cmd:aweight}s are allowed; see {help weight}.


{title:Description}

{pstd}
{cmd:graph} {cmd:twoway} {cmd:kdensity} plots a kernel density estimate for
{it:varname} using {cmd:graph} {cmd:twoway} {cmd:line}.


{title:Options}

{phang}
{cmd:width(}{it:#}{cmd:)}
and
{cmd:epan},
{cmd:epan2}, ...,
{cmd:triangle}
    specify how the kernel density estimate is to be obtained and are in fact
    the same options as those specified with the command {cmd:kdensity}; see 
    {helpb kdensity}.

{pmore}
    {cmd:width(}{it:#}{cmd:)}
    specifies the smoothing parameter.

{pmore}
    {cmd:epan},
    {cmd:epan2}, ...,
    {cmd:triangle}
    are alternatives and
    specify the kernel-weight function to be used.
    {cmd:epan} (Epanechnikov) is the default.

{pmore}
    See {helpb kdensity} for more information about these options.

{pmore}
    All the other {cmd:graph} {cmd:twoway} {cmd:kdensity} options modify how
    the result is displayed, not how it is obtained.

{phang}
{cmd:range(}{it:#} {it:#}{cmd:)}
and
{opth range(varname)}
    specify the range of values at which the kernel density estimates are
    to be plotted.  The default is {bind:{cmd:range(}{it:m} {it:M}{cmd:)}},
    where {it:m} and {it:M} are the minimum and maximum of the {it:varname}
    specified on the {cmd:graph} {cmd:twoway} {cmd:kdensity} command.

{pmore}
    {cmd:range(}{it:#} {it:#}{cmd:)}
    specifies a pair of numbers to be used as the minimum and maximum.

{pmore}
    {cmd:range(}{it:varname}{cmd:)}
    specifies another variable for which its minimum and maximum are to be
    used.

{phang}
{cmd:n(}{it:#}{cmd:)}
    specifies the number of points at which the estimate is evaluated.  The
    default is {cmd:n(300)}.

{phang}
{cmd:area(}{it:#}{cmd:)}
    specifies a multiplier by which the density estimates are adjusted before
    being plotted.  The default is {cmd:area(1)}.  {cmd:area()} is useful when
    overlaying a density estimate on top of a histogram that is itself not
    scaled as a density.  For instance, if you wished to scale the density
    estimate as a frequency, {cmd:area()} would be specified as the total
    number of nonmissing observations.

{phang}
{cmd:horizontal}
    specifies that the result be plotted horizontally (i.e, reflected along
    the identity line).

{phang}
{it:cline_options}
     specify how the density line is rendered and its appearance;
     see {it:{help cline_options}}.

INCLUDE help gr_axlnkf

INCLUDE help gr_twoptf


{title:Remarks}

{pstd}
{cmd:graph} {cmd:twoway} {cmd:kdensity} {it:varname} uses the {cmd:kdensity}
command to obtain an estimate of the density of {it:varname} and uses
{cmd:graph} {cmd:twoway} {cmd:line} to plot the result.

{pstd}
Remarks are presented under the headings

	{help twoway kdensity##remarks1:Typical use}
	{help twoway kdensity##remarks1:Use with by}


{marker remarks1}{...}
{title:Typical use}

{pstd}
The density estimate is often graphed on top of the histogram:

	{cmd:. sysuse lifeexp, clear}

	{cmd:. tw histogram lexp, color(*.5) || kdensity lexp}
	  {it:({stata "gr_example lifeexp: tw histogram lexp, color(*.5) || kdensity lexp":click to run})}
{* graph gtkden1}{...}

{* fill areas, dimming and brightening}{...}
{* index colors, dimming and brightening}{...}
{* index color() tt option}{...}
{* index color intensity adjustment}{...}
{* index intensity, color, adjustment}{...}
{pstd}
Notice the use of {cmd:graph} {cmd:twoway} {cmd:histogram}'s
{cmd:color(*.5)} option to dim the bars and make the line stand out;
see {it:{help colorstyle}}.

{pstd}
Notice also the {it:y} and {it:x} axis titles:  "Density/kdensity lexp" and
"Life expectancy at birth/x".  The "kdensity lexp" and "x" were contributed by
the {cmd:twoway} {cmd:kdensity}.  When you overlay graphs, you nearly always
need to respecify the axis titles using the {it:axis_title_options}
{cmd:ytitle()} and {cmd:xtitle()}; see {it:{help axis_title_options}}.


{marker remarks2}{...}
{title:Use with by}

{pstd}
{cmd:graph} {cmd:twoway} {cmd:kdensity} may be used with {cmd:by()}:

	{cmd:. sysuse lifeexp, clear}

{phang2}
	{cmd:. tw histogram lexp, color(*.5) || kdensity lexp ||, by(region)}
{p_end}
	  {it:({stata "gr_example lifeexp: tw histogram lexp, color(*.5) || kdensity lexp ||, by(region)":click to run})}
{* graph gtkden2}{...}


{title:Also see}

{psee}
Manual:  {bf:[G] graph twoway kdensity}

{psee}
Online:  
{helpb kdensity};
{helpb twoway histogram}
{p_end}