| Linear mixed effects parameters | |
|---|---|
| Fixed Effects formula | shannon ~ month * delivery * diet * sex |
| Metric | shannon |
| Group column | [delivery, diet, sex] |
| State column | month |
| Individual ID column | studyid |
| Random effects | None |
Model summary
| model summary | |
|---|---|
| Model: | MixedLM |
| No. Observations: | 875 |
| No. Groups: | 43 |
| Min. group size: | 2 |
| Max. group size: | 34 |
| Mean group size: | 20.3 |
| Dependent Variable: | shannon |
| Method: | REML |
| Scale: | 0.4564 |
| Log-Likelihood: | -961.5178 |
| Converged: | Yes |
Download summary as tsv
Model results
| Coef. | Std.Err. | z | P>|z| | [0.025 | 0.975] | |
|---|---|---|---|---|---|---|
| Intercept | 2.161 | 0.224 | 9.662 | 0.000 | 1.722 | 2.599 |
| delivery[T.Vaginal] | 0.583 | 0.298 | 1.958 | 0.050 | -0.001 | 1.167 |
| diet[T.fd] | 0.854 | 0.280 | 3.045 | 0.002 | 0.304 | 1.404 |
| sex[T.Male] | 0.317 | 0.267 | 1.187 | 0.235 | -0.206 | 0.841 |
| delivery[T.Vaginal]:diet[T.fd] | -1.066 | 0.528 | -2.019 | 0.044 | -2.101 | -0.031 |
| delivery[T.Vaginal]:sex[T.Male] | -0.999 | 0.347 | -2.882 | 0.004 | -1.679 | -0.320 |
| diet[T.fd]:sex[T.Male] | -0.566 | 0.432 | -1.310 | 0.190 | -1.412 | 0.281 |
| delivery[T.Vaginal]:diet[T.fd]:sex[T.Male] | 1.302 | 0.671 | 1.940 | 0.052 | -0.013 | 2.617 |
| month | 0.095 | 0.010 | 9.573 | 0.000 | 0.075 | 0.114 |
| month:delivery[T.Vaginal] | -0.008 | 0.013 | -0.623 | 0.533 | -0.034 | 0.017 |
| month:diet[T.fd] | -0.050 | 0.013 | -3.735 | 0.000 | -0.076 | -0.024 |
| month:sex[T.Male] | -0.011 | 0.013 | -0.846 | 0.397 | -0.037 | 0.014 |
| month:delivery[T.Vaginal]:diet[T.fd] | 0.054 | 0.028 | 1.900 | 0.057 | -0.002 | 0.110 |
| month:delivery[T.Vaginal]:sex[T.Male] | 0.053 | 0.016 | 3.216 | 0.001 | 0.021 | 0.085 |
| month:diet[T.fd]:sex[T.Male] | 0.025 | 0.029 | 0.863 | 0.388 | -0.032 | 0.081 |
| month:delivery[T.Vaginal]:diet[T.fd]:sex[T.Male] | -0.046 | 0.040 | -1.128 | 0.259 | -0.125 | 0.034 |
| Group Var | 0.103 | 0.047 |
Download results as tsv
Regression scatterplots
Projected Residuals
Fit values for metric vs. observation error (residuals). Residuals should be roughly zero-centered and normal. Uncentered, systematically high or low, and autocorrelated values could indicate a poor model.
