When I run mr(dat), the warning message goes "summary may be unreliable" and "number of items to replace is not a multiple of replacement length"

[Spasmodic114]

The warning message was:

1. In summary.lm(mod) : essentially perfect fit: summary may be unreliable
2. In beta[length(beta) + 1] <- densityIV$x[densityIV$y == max(densityIV$y)] :
   number of items to replace is not a multiple of replacement length

[remlapmot]

Your choice of outcome and/or exposure and/or instruments is clearly not good because you have a regression model with perfect fit, i.e., you are regressing two variables which are very highly correlated, e.g., something like

x <- 1:10
y <- 1:10
summary(lm(y ~ x))
#> Warning in summary.lm(lm(y ~ x)): essentially perfect fit: summary may be
#> unreliable
#> 
#> Call:
#> lm(formula = y ~ x)
#> 
#> Residuals:
#>        Min         1Q     Median         3Q        Max 
#> -5.356e-16 -3.618e-17  4.623e-17  1.685e-16  2.057e-16 
#> 
#> Coefficients:
#>              Estimate Std. Error   t value Pr(>|t|)    
#> (Intercept) 0.000e+00  1.675e-16 0.000e+00        1    
#> x           1.000e+00  2.699e-17 3.704e+16   <2e-16 ***
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> 
#> Residual standard error: 2.452e-16 on 8 degrees of freedom
#> Multiple R-squared:      1,  Adjusted R-squared:      1 
#> F-statistic: 1.372e+33 on 1 and 8 DF,  p-value: < 2.2e-16

^{Created on 2025-03-24 with reprex v2.1.1}

So you need to change your choice of outcome/exposure/instruments.