R Language Learning notes: simple regression analysis

Fitbit <-read.csv ("Fitbit.csv")

Date cal Step Dist floor sit inactive walk run
1 January 1, 2014 2496 12803 9.14 15 1161 123 98 58
2 January 2, 2014 2513 12955 9.47 12 1188 112 67 73
3 January 3, 2014 2228 8768 6.31 16 1234 97 72 37
4 January 4, 2014 2336 8971 6.41 16 1140 174 113 13
5 January 5, 2014 2508 12078 9.05 8 1153 130 108 49
6 January 6, 2014 2763 15880 11.36 20 1061 177 141 61
7 January 7, 2014 2453 10895 7.78 12 1145 156 99 40
8 January 8, 2014 2449 11692 8.44 13 1164 121 118 37

Attach (Fitbit) #可以直接用step表示fitbit $step.

Plot (step, Dist)

Looking at the correlation between step and dist, you can see that the value is almost equal to 1, indicating that the step number is strongly correlated with the distance.

Cor (step, Dist)
[1] 0.9998078

For a simple linear regression:

Dist.step <-LM (dist ~ step, Fitbit)

Display the value of Dist.step to see:

Call : lm (formula = Dist ~ step, data = Fitbit)

Coefficients:
(Intercept) Step
-0.0059966 0.0007171

Description Fit formula: Dist = -0.0059966 + 0.0007171 * Step

You can use the Abline function to draw a fitting line:

Abline (Dist.step)

There is also a more cow car bag, you can draw a more powerful graphics at once:

Library (CAR)

Scatterplot (dist ~ step, data = Fitbit)

Next to the coordinate system there is a box-line diagram, which automatically draws the fitted line.

You can look at the relationship between calorie consumption and step count:

Scatterplot (cal ~ step, data = Fitbit, Spread=false)

You can see that this function automatically fits a straight line and a curve.

R Language Learning notes: simple regression analysis