Skip to content

Plotting model over data

2 messages · Kare Edvardsen, PIKAL Petr

#
Dear R-list.

How would I plot model result over actual data?

Here are the lme results:



Linear mixed-effects model fit by REML
  Data: D-set
        AIC     BIC    logLik
   884.3326 911.428 -432.1663

Random effects:
  Formula: ~C | ID
  Structure: General positive-definite, Log-Cholesky parametrization
             StdDev     Corr
(Intercept) 13.7293823 (Intr)
C            0.7425786 -0.544
Residual     6.7688776

Fixed effects: Y ~ A * C + B * C
                 Value Std.Error DF   t-value p-value
(Intercept) 29.662389  7.979584 98  3.717285  0.0003
A            0.185470  0.102510 13  1.809294  0.0936
C           -0.741605  0.692078 98 -1.071563  0.2865
B            0.006185  0.002265 98  2.730804  0.0075
A:C         -0.003325  0.008243 98 -0.403405  0.6875
C:B         -0.000679  0.000278 98 -2.444981  0.0163
  Correlation:
          (Intr) A      C      B      A:C
A        -0.865
C        -0.540  0.437
B        -0.138 -0.009  0.104
A:C       0.473 -0.559 -0.770  0.030
C:B       0.152  0.013 -0.218 -0.958 -0.042

Standardized Within-Group Residuals:
         Min          Q1         Med          Q3         Max
-3.55799126 -0.53364258  0.01324915  0.52255745  2.68609309

Number of Observations: 117
Number of Groups: 15
#
Hi

?predict
over
?order #(ed)
data
could be one way.

HTH
Petr
On 14 Dec 2005 at 13:36, Kare Edvardsen wrote:
Date sent:      	Wed, 14 Dec 2005 13:36:42 +0100
From:           	Kare Edvardsen <ked at nilu.no>
To:             	R-help <r-help at stat.math.ethz.ch>
Subject:        	[R] Plotting model over data
Petr Pikal
petr.pikal at precheza.cz