Hey there! Let's work through your coefplot questions step by step:

1. Fixing the command headings is unrecognized Error #

The headings option is a relatively new feature in coefplot (added in version 1.1.0 or later). If you're seeing this error, your installed version of coefplot is outdated. Here are two ways to fix it:

Option 1: Update coefplot #

Run this command in Stata to install the latest version from the Boston College archive:

ssc install coefplot, replace

Once updated, your original code with the headings option should run without issues.

Option 2: Workaround for Older coefplot Versions #

If you can't update, you can mimic the section header effect using coeflabels() and hiding marker symbols for header lines. Here's how to adjust your example:

sysuse auto, clear
keep if rep78>=3
regress mpg headroom i.rep##i.foreign
coefplot, xline(0) omitted baselevels drop(_cons) ///
coeflabels(3.rep78 = "{bf:Repair Record}" 3.rep78 = "Repair Score 3" ///
           0.foreign = "{bf:Car Type}" 0.foreign = "Domestic" ///
           3.rep78#0.foreign = "{bf:Interaction Effects}" 3.rep78#0.foreign = "Repair 3 × Domestic") ///
marker(1, ms(none) mlabcolor(black)) marker(3, ms(none) mlabcolor(black)) marker(5, ms(none) mlabcolor(black))

This uses custom coefficient labels for section headers and hides their markers to make them stand out like the headings option would.

2. Setting the Y-axis Title (Overall Axis Label) #

To set the main label for the entire Y-axis, use the ytitle() option in your coefplot command. Here's how to add it to your initial regression code:

sysuse auto, clear
regress price mpg trunk length turn if foreign==0
iestimates store Option1
regress price mpg trunk length turn if foreign==1
iestimates store Option2
coefplot Option1 Option2, drop(_cons) xline(1) ///
ytitle("Regression Variables")  // This defines the Y-axis title

3. Renaming Individual Y-axis Variables (Coefficient Labels) #

To rename specific variables on the Y-axis (like changing "Mileage (mpg)" to "Mileage"), you have two easy options:

Option 1: Use coeflabels() Directly in coefplot #

Map each coefficient name to your custom label within the command:

sysuse auto, clear
regress price mpg trunk length turn if foreign==0
iestimates store Option1
regress price mpg trunk length turn if foreign==1
iestimates store Option2
coefplot Option1 Option2, drop(_cons) xline(1) ///
ytitle("Regression Variables") ///
coeflabels(mpg = "Mileage" ///
           trunk = "Trunk Space" ///
           length = "Car Length" ///
           turn = "Turn Radius")

Option 2: Set Variable Labels Before Regression #

Define variable labels in Stata first, and coefplot will automatically use them:

sysuse auto, clear
label variable mpg "Mileage"
label variable trunk "Trunk Space"
label variable length "Car Length"
label variable turn "Turn Radius"

regress price mpg trunk length turn if foreign==0
iestimates store Option1
regress price mpg trunk length turn if foreign==1
iestimates store Option2

coefplot Option1 Option2, drop(_cons) xline(1) ytitle("Regression Variables")

This avoids repeating labels in the coefplot command and keeps your code cleaner.

内容的提问来源于stack exchange，提问作者Quinn