在 Plotly 中将回归平面添加到 3d 散点图 [英] Add Regression Plane to 3d Scatter Plot in Plotly

问题描述

我希望利用 Plotly 中的强大功能，但我很难弄清楚如何将回归平面添加到 3d 散点图中.这是一个如何开始使用 3d 绘图的示例，有人知道如何进行下一步并添加平面吗?

图书馆(情节)数据(虹膜)iris_plot <- plot_ly(my_df,x = 萼片长度，y = Sepal.Width,z =花瓣.长度，类型=scatter3d"，模式=标记")花瓣 lm <- lm(花瓣长度 ~ 0 + 萼片长度 + 萼片宽度，数据=虹膜)

解决方案

您需要根据 lm 调用创建的预测对象对点进行采样.这将创建一个类似于

I am looking to take advantage of the awesome features in Plotly but I am having a hard time figuring out how to add a regression plane to a 3d scatter plot. Here is an example of how to get started with the 3d plot, does anyone know how to take it the next step and add the plane?

library(plotly)
data(iris)


iris_plot <- plot_ly(my_df, 
                x = Sepal.Length, 
                y = Sepal.Width, 
                z = Petal.Length, 
                type = "scatter3d", 
                mode = "markers")

petal_lm <- lm(Petal.Length ~ 0 + Sepal.Length + Sepal.Width, 
               data = iris)

解决方案

You need to sample the points based on the predict object created from your lm call. This creates a surface similar to the volcano object which you can then add to your plot.

library(plotly)
library(reshape2)

#load data

my_df <- iris
petal_lm <- lm(Petal.Length ~ 0 + Sepal.Length + Sepal.Width,data = my_df)

The following sets up the extent of our surface. I chose to sample every 0.05 points, and use the extent of the data set as my limits. Can easily be modified here.

#Graph Resolution (more important for more complex shapes)
graph_reso <- 0.05

#Setup Axis
axis_x <- seq(min(my_df$Sepal.Length), max(my_df$Sepal.Length), by = graph_reso)
axis_y <- seq(min(my_df$Sepal.Width), max(my_df$Sepal.Width), by = graph_reso)

#Sample points
petal_lm_surface <- expand.grid(Sepal.Length = axis_x,Sepal.Width = axis_y,KEEP.OUT.ATTRS = F)
petal_lm_surface$Petal.Length <- predict.lm(petal_lm, newdata = petal_lm_surface)
petal_lm_surface <- acast(petal_lm_surface, Sepal.Width ~ Sepal.Length, value.var = "Petal.Length") #y ~ x

At this point, we have petal_lm_surface, which has the z value for every x and y we want to graph. Now we just need to create the base graph (the points), adding color and text for each species:

hcolors=c("red","blue","green")[my_df$Species]
iris_plot <- plot_ly(my_df, 
                     x = ~Sepal.Length, 
                     y = ~Sepal.Width, 
                     z = ~Petal.Length,
                     text = ~Species, # EDIT: ~ added
                     type = "scatter3d", 
                     mode = "markers",
                     marker = list(color = hcolors))

and then add the surface:

iris_plot <- add_trace(p = iris_plot,
                       z = petal_lm_surface,
                       x = axis_x,
                       y = axis_y,
                       type = "surface")

iris_plot

这篇关于在 Plotly 中将回归平面添加到 3d 散点图的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！