尚未检测到更新。GitHub 发布版本后会显示在这里。
个人 GitHub 账号下公开仓库的所有者可以登录并验证所有权,自行管理介绍,无需管理员批准。
登录并生成验证文件,在 30 分钟内将其提交到此仓库的默认分支。其他分支或分叉仓库中的文件无效。然后在下方检查文件,即可无需管理员批准直接管理项目。请保留文件以维持管理权限。
如需登录,将前往登录页面。登录后请回到此处,再次点击所选的验证按钮。
在项目文件夹中运行 /nomorevibe verify,即可验证所有权并自行管理介绍。
一分钟完成验证制作者的介绍仅有上面的一句话。 更多详情请阅读仓库 README。 仓库 README ↗
正在加载完整 README。
查看完整 README ↗An Excel add-in that runs forward selection, backward elimination and stepwise regression, then writes a report in the same layout as Excel's Data Analysis > Regression tool (with Standard Error shown as RMSE). Variables enter and leave on |T-Stat| ≥ 2 by default, or on p-values if you prefer.
Install page: https://cranksrust.github.io/stepwise-regression-excel/
The add-in is not in the Office Add-ins store yet, so it is installed from a small file called HOUMBA-28-Stepwise-Regression.xml (Microsoft calls this a manifest: it tells Excel the add-in's name, button and web address). This takes about two minutes and only needs to be done once per browser or computer.
HOUMBA-28-Stepwise-Regression.xml somewhere easy to find, such as Downloads.Excel for the web (recommended, works for everyone)
HOUMBA-28-Stepwise-Regression.xml, and click Upload.Excel for Mac
~/Library/Containers/com.microsoft.Excel/Data/Documents/wef and press Return. If it says the folder does not exist, go to ~/Library/Containers/com.microsoft.Excel/Data/Documents and create a folder named wef.HOUMBA-28-Stepwise-Regression.xml into the wef folder.Excel for Windows (desktop)
Go to Home > Add-ins > More Add-ins > My Add-ins. If Upload My Add-in is shown, follow steps 4 to 6 above. If it is not, use Excel for the web.
Download home-prices.xlsx: 60 fictional home sales with eight candidate predictors. Use Y A1:A61 and X B1:I61 on the Home Prices sheet, then run Forward, Backward and Stepwise and compare the steps.
| Step | Forward | Backward | Stepwise |
|---|---|---|---|
| Start | intercept only | all 8 | intercept only |
| 1 | + TAX_ASSESSED | − GARAGE | + TAX_ASSESSED |
| 2 | + AGE | − BATHROOMS | + AGE |
| 3 | + SQFT | − BEDROOMS | + SQFT |
| 4 | + MILES_DOWNTOWN | − TAX_ASSESSED | + MILES_DOWNTOWN |
| 5 | + SCHOOL_RATING | − TAX_ASSESSED | |
| 6 | + SCHOOL_RATING | ||
| Final R Square | 0.9082 (5 terms) | 0.9070 (4 terms) | 0.9070 (4 terms) |
TAX_ASSESSED looks like the best single predictor, but it only echoes the real drivers. Stepwise drops it once they are in the model (its T-Stat falls to 1.58). Forward selection never re-checks, so it ends with TAX_ASSESSED still in the model at a T-Stat of −0.85.
| Problem | Fix |
|---|---|
| No Add-ins button | The window is too narrow. Check under ⋯ on the Home ribbon or the Insert tab, or widen the window. |
| Upload My Add-in is missing or greyed out | Your school or work account may block custom add-ins. Sign in to office.com with a personal Microsoft account and try again. |
| The button disappeared | In Excel for the web, uploaded add-ins are remembered per browser. Upload the add-in file again. |
| "Every X column must start and end on the same rows as Y" | Select the same rows for Y and X, for example A4:A48 and B4:F48. |
| "The model with every predictor cannot be estimated" | Two X columns carry the same information (for example one is a fixed number minus the other). Remove one, or use Forward or Stepwise. |
More help: https://cranksrust.github.io/stepwise-regression-excel/support.html
Built by the University of Texas at Austin's MBA at Houston Class of '28 for Professor Tom Sager's BA 286T Statistics course, Fall 2026.