<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Software | Jesper N. Wulff</title><link>https://jespernwulff.github.io/software/</link><atom:link href="https://jespernwulff.github.io/software/index.xml" rel="self" type="application/rss+xml"/><description>Software</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><image><url>https://jespernwulff.github.io/media/icon_hu_c98bc1085d22e9da.png</url><title>Software</title><link>https://jespernwulff.github.io/software/</link></image><item><title>alphaN</title><link>https://jespernwulff.github.io/software/alphan/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://jespernwulff.github.io/software/alphan/</guid><description>&lt;p&gt;&lt;code&gt;alphaN&lt;/code&gt; implements the Bayesian–frequentist compromise from Wulff &amp;amp; Taylor (2024): it returns a sample-size-dependent significance threshold derived from Bayes factors, so that the alpha level shrinks as the sample grows. An interactive web app built on the same method lets users explore thresholds without writing code.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;install.packages(&amp;quot;alphaN&amp;quot;)
library(alphaN)
alphaN(n = 500)
&lt;/code&gt;&lt;/pre&gt;</description></item><item><title>biokNN</title><link>https://jespernwulff.github.io/software/bioknn/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://jespernwulff.github.io/software/bioknn/</guid><description>&lt;p&gt;&lt;code&gt;biokNN&lt;/code&gt; provides a bi-objective k-nearest-neighbours imputation method for multilevel data, balancing global structure and cluster-level structure when filling in missing values. It accompanies Cubillos, Wøhlk &amp;amp; Wulff (2022).&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;install.packages(&amp;quot;biokNN&amp;quot;)
&lt;/code&gt;&lt;/pre&gt;</description></item><item><title>bootmakr</title><link>https://jespernwulff.github.io/software/bootmakr/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://jespernwulff.github.io/software/bootmakr/</guid><description>&lt;p&gt;&lt;code&gt;bootmakr&lt;/code&gt; adds bootstrap inference to &lt;a href="https://carloscinelli.com/sensemakr/" target="_blank" rel="noopener"&gt;sensemakr&lt;/a&gt;-style sensitivity analysis, so that sensitivity statistics come with resampling-based uncertainty rather than point estimates alone. A companion Stata implementation is available as &lt;a href="../bootmakr-stata/"&gt;bootmakr for Stata&lt;/a&gt;.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;# install.packages(&amp;quot;remotes&amp;quot;)
remotes::install_github(&amp;quot;jespernwulff/bootmakr&amp;quot;)
&lt;/code&gt;&lt;/pre&gt;</description></item><item><title>bootmakr for Stata</title><link>https://jespernwulff.github.io/software/bootmakr-stata/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://jespernwulff.github.io/software/bootmakr-stata/</guid><description>&lt;p&gt;The Stata implementation of &lt;a href="../bootmakr/"&gt;bootmakr&lt;/a&gt;: bootstrap inference for &lt;code&gt;sensemakr&lt;/code&gt; sensitivity analysis, with support for clustered bootstrap, benchmark covariates, multiple &lt;code&gt;kd&lt;/code&gt; values, and plots.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;net install bootmakr, from(&amp;quot;https://raw.githubusercontent.com/jespernwulff/bootmakr-stata/main/&amp;quot;)
&lt;/code&gt;&lt;/pre&gt;</description></item><item><title>ginteff</title><link>https://jespernwulff.github.io/software/ginteff/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://jespernwulff.github.io/software/ginteff/</guid><description>&lt;p&gt;&lt;code&gt;ginteff&lt;/code&gt; is an R port of the Stata &lt;a href="https://doi.org/10.1177/1536867X231175253" target="_blank" rel="noopener"&gt;ginteff&lt;/a&gt; command by Marius Radean (&lt;em&gt;The Stata Journal&lt;/em&gt;, 2023). It computes two- and three-way interaction effects — via partial derivatives or first differences — for fitted regression models, with delta-method standard errors.&lt;/p&gt;
&lt;p&gt;Built as a thin wrapper around &lt;a href="https://marginaleffects.com" target="_blank" rel="noopener"&gt;marginaleffects&lt;/a&gt;, it supports arbitrary variance–covariance specifications (&lt;code&gt;&amp;quot;HC3&amp;quot;&lt;/code&gt;, clustered, or a user-supplied sandwich matrix), which propagate through to the final interaction-effect standard errors.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;# install.packages(&amp;quot;remotes&amp;quot;)
remotes::install_github(&amp;quot;jespernwulff/ginteff&amp;quot;)
&lt;/code&gt;&lt;/pre&gt;</description></item></channel></rss>