line
This command draws straight or curved lines. It can be used to graph the results of a regression.
Needs to be rewritten
Example
. drop if earndad > 10000
(2 observations deleted)
. reg award earndad, r
Regression with robust standard errors Number of obs = 279
F( 1, 277) = 43.07
Prob > F = 0.0000
R-squared = 0.2979
Root MSE = 200.91
------------------------------------------------------------------------------
| Robust
award | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
earndad | .0770272 .0117365 6.56 0.000 .0539231 .1001313
_cons | 265.7285 14.62653 18.17 0.000 236.9352 294.5218
------------------------------------------------------------------------------
. predict award_hat
(option xb assumed; fitted values)
. line award_hat earndad, sort
Options
- The option , sort is needed to make Stata connect the points in order, and draw a line, not abstract art.
Notes
- A polynomial or log curve can be drawn in just the same way:
. gen ed_2 = earndad * earndad . reg award earndad ed_2, r Regression with robust standard errors Number of obs = 279 F( 2, 276) = 36.76 Prob > F = 0.0000 R-squared = 0.3042 Root MSE = 200.37 ------------------------------------------------------------------------------ | Robust award | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- earndad | .1015935 .0237884 4.27 0.000 .0547637 .1484233 ed_2 | -4.05e-06 4.66e-06 -0.87 0.386 -.0000132 5.13e-06 _cons | 253.5037 14.97709 16.93 0.000 224.0199 282.9875 ------------------------------------------------------------------------------ . predict a_hat (option xb assumed; fitted values) . line a_hat earndad, sort
- This command may be combined with scatter to overlay the regression line on the data, see graphing the regression line.
- See also scatter and histogram.