Skip to main content

Table 5 Akaike information criterion (AIC) of models at first stage (M1, , M9) by year and location (L) for grain dry matter yield (Y)

From: The importance of phenotypic data analysis for genomic prediction - a case study comparing different spatial models in rye

Year

L

M1

M2

M3

M4

M5

M6

M7

M8

M9

2009

G-L1

101.7

84.3

45.5

47.2

20.4

6.9

45.6

0

1.7

2009

G-L2

83.1

67.5

50.9

38.5

31.4

20.7

40.7

0.5

0

2009

G-L3

45.7

30.4

41.5

31.1

40.1

26.9

31.2

1.0

0

2009

G-L4

125.0

19.1

125.1

114.9

90.3

19.6

115.5

65.0

0

2009

G-L5

29.1

8.0

18.1

24.5

15.3

1.2

12.3

0

2009

G-L6

51.6

47.6

37.7

29.5

41.7

35.4

29.4

0

1.2

2009

G-L7

81.5

56.1

55.3

62.8

36.5

11.0

55.5

5.1

0

2009

P-L1

126.4

115.6

121.6

116.3

109.5

108.8

116.2

0

1.9

2009

P-L2

62.3

45.4

62.4

54.6

57.3

47.2

54.9

1.5

0

2009

P-L3

120.9

65.9

116.1

105.5

99.7

49.6

105.5

17.3

0

2009

P-L4

145.9

98.6

132.8

126.4

126.4

80.1

126.4

0.4

0

2010

G-L1

35.5

4.9

35.6

31.5

12.3

0

32.0

12.3

1.8

2010

G-L2

25.0

7.2

27.0

21.7

29.7

11.9

19.7

0

-3.2

2010

G-L4

141.4

74.2

128.7

117.1

130.2

57.4

118.4

5.0

0

2010

G-L5

21.6

0

23.4

22.9

21.9

3.3

22.9

22.1

2.8

2010

G-L6

80.9

60.0

72.8

59.8

55.4

41.5

61.1

0

0.6

2010

G-L7

69.5

22.3

56.2

47.8

37.2

23.6

48.1

2.6

0

2010

G-L8

40.8

24.7

32.1

22.6

27.7

19.6

23.1

0

1.4

2010

P-L1

38.8

5.7

38.8

38.8

39.4

9.4

40.8

39.1

0

2010

P-L2

40.0

0.7

41.6

36.1

39.8

4.1

36.9

4.3

0

2010

P-L3

66.4

0

68.4

67.2

69.5

3.7

70.4

71.5

5.7

2010

P-L4

95.0

80.4

90.5

79.1

87.0

66.7

79.4

0

3.2

Counts

0

2

0

0

0

1

0

7

12

 

0%

9%

0%

0%

0%

5%

0.00

32%

55%

2012

G-L4

35.3

0

35.3

36.2

26.0

0.6

35.3

24.2

2012

G-L5

66.3

2.6

67.0

66.3

42.1

5.9

21.5

0

2012

G-L6

148.4

131.4

93.8

93.7

18.7

18.7

89.9

0

0

2012

G-L7

38.3

4.5

40.3

38.3

36.3

0

42.3

1.9

2012

G-L8

45.3

39.8

37.7

33.5

35.6

37.3

33.9

1.9

0

2012

G-L9

402.3

321.5

200.9

181.7

81.9

81.9

191.6

0

0

2012

G-L10

39.7

0

41.5

41.4

22.1

3.5

43.5

6.7

1.1

2012

G-L11

18.0

0

19.7

18.0

8.4

1.2

21.6

3.7

2012

P-L1

189.5

168.8

158.9

148.9

146.3

137.8

149.1

0

1.7

2012

P-L2

127.4

49.3

129.1

122.6

129.7

49.9

123.9

5.9

0

2012

P-L3

107.8

55.3

103.1

95.0

101.0

49.3

96.1

7.9

0

2012

P-L4

226.3

0.2

226.3

222.1

226.3

0

226.3

226.3

2.0

2012

P-L5

13.2

0

13.2

13.2

11.9

1.5

13.2

13.9

3.5

2012

P-L6

79.0

54.8

70.4

66.9

65.8

37.9

67.0

0

1.7

Counts

0

4

0

0

0

2

0

4

6

 

0%

29%

0%

0%

0%

14%

0%

29%

43%

  1. Table shows ΔAIC relative to the best model. Boldfaced entries in the table indicate best model (fit) within location. Empty cells (–) correspond to locations where the model did not converge. In italics, we report the models that converged but the Hessian matrix was not positive definite.