Food environments that were classified as having a lower proportion of healthier options cluster into two sub-types, one dominated by convenience stores and the other by restaurants and fast food establishments.
In addition, we use the Stata option "cluster", i.e. cluster(country), which takes into account the nested structure of the data (i.e. the fact that individuals are nested within countries) and adjusts the standard errors accordingly.
The GeneMaths option Cluster Plot was used to plot the logarithm of the cluster size versus the cluster similarity.
Using the GeneMaths option Cluster Plot (see Methods section for details) provided a recommended branch cutoff level of 88% cluster similarity.
We used the option "cluster groups of individuals" to run the program with default conditions.
bAdjusted odds rations (AOR) and p values are based on logistic regression models with cluster option (cluster variable = outlet where data came from).
Statistical analyses were conducted on an intention-to-treat basis using Stata 9 (Stata Statistical Software, 2005, College Station, TX, USA) with the option cluster to adjust the standard errors for intra-group correlation.
No effects on the perceived degree of healthiness of the last meal consumed in the restaurant were observed.> -wrap-foot> aAdjusted coefficients (B) and p values are based on multiple linear regression models with cluster option (cluster variable = outlet where data came from).
In this section, users have three options: Calculate clusters of points, Detect outliers and clusters, and Predict trends and volatility.
The main features include many different options for clustering of compounds as well as commenting on clusters, cluster visualization, prioritization and interactive reclustering.
