https://communities.sas.com/t5/Statistical-Procedures/How-to-obtain-density-estimates-with-chosen-kernel-and-number-of/m-p/739418#M35928 <P>Moved to <EM>statistics</EM> community</P> Thu, 06 May 2021 07:47:29 GMT ChrisNZ 2021-05-06T07:47:29Z https://communities.sas.com/t5/Statistical-Procedures/How-to-obtain-density-estimates-with-chosen-kernel-and-number-of/m-p/737541#M35927 <P>Dear community,</P><P>&nbsp;</P><P>I would please like to ask for your help concerning the following issue.</P><P>&nbsp;</P><P>How do you obtain density estimates with chosen kernel and number of grid points?</P><P>Using <FONT face="courier new,courier">proc univariate</FONT>, together with the <FONT face="courier new,courier">histogram</FONT>-statement, you <A href="https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/procstat/procstat_univariate_syntax09.htm#procstat.univariate.hist_k" target="_self">can</A> choose between normal, quadratic and triangular kernels; however, regardless of whether you use&nbsp;<FONT face="courier new,courier">outkernel = </FONT>or <FONT face="courier new,courier">ods output histogram =</FONT>, the number of grid points is fixed, as far as I see.</P><P><FONT face="courier new,courier">proc kde</FONT>, on the other hand, allows to specify <FONT face="courier new,courier">NGRID=</FONT>, but, you seem to be restricted to a normal kernel.</P><P>&nbsp;</P><P>Yours sincerely,</P><P>Mario</P> Wed, 28 Apr 2021 11:30:06 GMT https://communities.sas.com/t5/Statistical-Procedures/How-to-obtain-density-estimates-with-chosen-kernel-and-number-of/m-p/737541#M35927 Sinistrum 2021-04-28T11:30:06Z https://communities.sas.com/t5/Statistical-Procedures/How-to-obtain-density-estimates-with-chosen-kernel-and-number-of/m-p/739418#M35928 <P>Moved to <EM>statistics</EM> community</P> Thu, 06 May 2021 07:47:29 GMT https://communities.sas.com/t5/Statistical-Procedures/How-to-obtain-density-estimates-with-chosen-kernel-and-number-of/m-p/739418#M35928 ChrisNZ 2021-05-06T07:47:29Z