Do remember you need to create your bins manually if you are using the scatterplot functionality from matplotlib as opposed to seaborn (depicted in the following example).Using the notion of facets for representing continous data in 3-DThe above plot clearly tells us that higher the residual_sugar levels and the alcohol content, lower is the fixed_acidity in the wine samples.Visualizing three-dimensional categorical data by introducing the notion of hue and facetsThe chart above clearly shows the frequency pertaining to each of the dimensions and you can see how easy and effective this can be in understanding relevant insights.Considering visualization for three mixed attributes, we can use the notion of hue for separating our groups in one of the categorical attributes while using conventional visualizations like scatter plots for visualizing two dimensions for numeric attributes.Visualizing mixed attributes in three-dimensions leveraging scatter plots and the concept of hueThus hue acts as a good separator for the categories or groups and while there is no or very weak correlation as observed above, we can still understand from these plots that sulphates are slightly higher for red wines as compared to white..Instead of a scatter plot, you can also use a kernel density plot to understand the data in three dimensions.Visualizing mixed attributes in three-dimensions leveraging kernel density plots and the concept of hueIt is quite evident and expected that red wine samples have higher sulphate levels as compared to white wines..You can also see the density concentrations based on the hue intensity.In case we are dealing with more than one categorical attribute in the three dimensions, we can use hue and one of the regular axes for visualizing data and use visualizations like box plots or violin plots to visualize the different groups of data.Visualizing mixed attributes in three-dimensions leveraging split violin plots and the concept of hueIn the figure above, we can see that in the 3-D visualization on the right hand plot, we have represented wine quality on the x-axis and wine_type as the hue..We can clearly see some interesting insights like volatile acidity is higher for red wines as compared to white wines.You can also consider using box plots for representing mixed attributes with more than one categorical variable in a similar way.Visualizing mixed attributes in three-dimensions leveraging box plots and the concept of hueWe can see that both for quality and quality_label attributes, the wine alcohol content increases with better quality..Also red wines tend to have a sightly higher median alcohol content as compared to white wines based on the quality class..However if we check the quality ratings, we can see that for lower rated wines (3 & 4), the white wine median alcohol content is greater than red wine samples..Otherwise red wines seem to have a slightly higher median alcohol content in general as compared to white wines.Visualizing data in Four Dimensions (4-D)Based on our discussion earlier, we leverage various components of the charts visualize multiple dimensions..One way to visualize data in four dimensions is to use depth and hue as specific data dimensions in a conventional plot like a scatter plot.Visualizing data in four-dimensions leveraging scatter plots and the concept of hue and depthThe wine_type attribute is denoted by the hue which is quite evident from the above plot..Also, while interpreting these visualizations start getting difficult due to the complex nature of the plots, you can still gather insights like fixed acidity is higher for red wines and residual sugar is higher for white wines..Of course if there were some association between alcohol and fixed acidity we might have seen a gradually increasing or decreasing plane of data points showing some trend.Is this effective?.Again, not really!.One strategy to make this better, is to keep a 2-D plot, but use hue and data point size as data dimensions.. More details