Search code examples
rggplot2regressionlinear-regression

How to get f(x)=kx+d form of linear regression line?

I have a data frame dt.Data with time data (values of this data frame are changing each day) and I'm plotting an correlation scatter plot and the regression line with ggplot(). The R code looks like this:

set.seed(123)

## Create data frame: ##
df.Data <- data.frame(date = seq(as.Date('2018-01-01'), by = '1 day', length.out = 1100),
                      DE = rnorm(1100, 2, 1), AT = rnorm(1100, 5, 2))
corPearson <- cor.test(x = df.Data$DE, y = df.Data$AT, method = "pearson")

df.Data$year <- format(as.Date(df.Data$date), '%Y')

p <- ggplot(data = df.Data, aes(x = DE, y = AT, group = 1)
      ) +
      geom_point(aes(color = year)) + 
      geom_smooth(method = "lm", se = FALSE, color = "#007d3c") +
      theme_classic() +
      theme(legend.position = "none") +
      theme(panel.background = element_blank()) +
      scale_colour_brewer(palette = 'Greens') + 
      xlab(product1) +
      ylab(product2) +
      ggtitle("Correlation Scatter Plot (Pearson)") +
      theme(plot.title = element_text(hjust = 0.5, face = "bold"))

# Correlation plot converting from ggplot to plotly: #
CorrelationPlot <- plotly::ggplotly(p, tooltip = "text")

The regression line is plotted with: geom_smooth(method = "lm", se = FALSE, color = "#007d3c").

The plot looks like this:

enter image description here

My question now is: How do I get the function of the regression line in the form f(x) = kx + d? I have already seen this question a few times in stackoverflow, but no answer there was complete or useful. Can someone help me?

EDIT: If I use this

reg <- lm(df.Data$AT ~ df.Data$DE)
summary(reg)

the output of the summary is:

enter image description here

where the d=5.07667 (red) and k=-0.03306 (blue)? Is this correct? How can I extract both values and construct a function like this: f(x)=kx+d=-0.3303x+5.07667??

I need this f(x) as an output of a valueBox() in a RShiny app.

Solution

  • You can use the lm() function:

    reg <- lm (df.Data$AT ~ df.Data$DE)
summary (reg)

    When you summarize you can see the intercept, which is your d and the slop which is you k. Let me know if this helps :)