如何使用逻辑回归制作概率数字? 我正在尝试用我的数据重制这个数字。 我有2个二进制因素(P处理和Embryo_presence_vs._absence)和1个连续值(final_size)。这是我的数据: print.data.frame(x [c(&...

问题描述 投票:0回答:1

I有2个二元因子(P处理和胚胎_PRESENCE_VS._ABSENCE)和1个连续值(final_size)。这是我的数据: print.data.frame(x[c("P_Treatment", "Embryo_Presence_vs._Absence", "Final_Size")]) P_Treatment Embryo_Presence_vs._Absence Final_Size 1 Intermediate 0 3.80 2 Intermediate 0 3.76 3 Intermediate 0 3.70 4 Intermediate 0 3.73 5 Intermediate 0 3.10 6 Intermediate 0 3.77 7 Intermediate 0 3.59 8 Intermediate 0 3.73 9 Intermediate 0 3.36 10 Intermediate 0 3.81 11 Intermediate 0 3.72 12 Intermediate 0 3.92 13 Intermediate 0 3.56 14 Intermediate 0 3.78 15 Intermediate 0 3.63 16 Intermediate 0 3.27 17 Intermediate 0 3.60 18 Intermediate 0 3.74 19 Intermediate 0 3.60 20 Intermediate 0 3.65 21 Intermediate 0 3.59 22 Intermediate 0 3.75 23 Intermediate 0 3.78 24 Intermediate 0 3.55 25 Intermediate 0 3.65 26 Intermediate 0 3.65 27 Low 0 3.91 28 Low 0 3.72 29 Low 0 3.77 30 Low 0 3.73 31 Low 0 3.65 32 Low 0 3.57 33 Low 0 3.82 34 Low 0 3.76 35 Low 0 3.88 36 Low 0 4.10 37 Low 0 3.71 38 Low 0 3.57 39 Low 0 3.77 40 Low 0 3.60 41 Low 0 3.70 42 Low 0 3.55 43 Low 0 4.00 44 Low 0 3.56 45 Low 0 3.71 46 Low 0 3.61 47 Low 0 3.63 48 Low 0 3.72 49 Low 0 3.80 50 Low 0 3.86 51 Low 0 3.08 52 Low 0 3.81 53 Low 0 3.73 54 Low 0 3.84 55 Low 0 3.76 56 Low 0 3.66 57 Low 0 3.70 58 Low 0 3.71 59 Low 0 3.60 60 Low 0 3.75 61 Low 0 3.74 62 Intermediate 1 3.89 63 Low 1 3.87 64 Intermediate 1 3.99 65 Intermediate 1 3.93 66 Intermediate 1 3.65 67 Intermediate 1 3.64 68 Intermediate 1 3.81 69 Low 1 3.67 70 Low 1 3.52 71 Intermediate 1 3.70 72 Low 1 3.83 73 Low 1 3.65 74 Low 1 3.94 75 Intermediate 1 3.75 76 Low 1 3.71 77 Intermediate 1 3.56 78 Intermediate 1 3.89 79 Low 1 3.72 80 Low 1 3.25 81 Intermediate 1 3.79 82 Intermediate 1 3.60 83 Intermediate 1 3.88 84 Intermediate 1 3.75 85 Intermediate 1 3.75 86 Low 1 3.58 87 Intermediate 1 3.75 88 Intermediate 1 3.65 89 Low 1 3.60 90 Intermediate 1 3.68 91 Low 1 3.65 92 Intermediate 1 3.88 93 Intermediate 1 3.77 94 Intermediate 1 3.63 95 Low 1 3.68 96 Intermediate 1 3.83 97 Intermediate 1 3.85 98 Low 1 3.81 99 Intermediate 1 3.56 100 Intermediate 1 3.70 101 Low 1 3.74 102 Low 1 3.65 103 Intermediate 1 3.69 104 Intermediate 1 3.89 105 Intermediate 1 3.70 106 Intermediate 1 3.70 107 Intermediate 1 3.76 108 Intermediate 1 3.66 109 Intermediate 1 3.78 110 Intermediate 1 4.22 111 Intermediate 1 3.68 112 Low 1 3.82 113 Low 1 3.82 114 Low 1 3.76 115 Low 1 3.91 116 Intermediate 1 3.56 117 Low 1 3.66 118 Intermediate 1 3.65 119 Low 1 3.61 120 Intermediate 1 3.65 121 Low 1 4.16 122 Intermediate 1 3.74 123 Intermediate 1 3.60 124 Low 1 3.50 125 Low 1 3.76 126 Low 1 3.85 127 Low 1 3.83 128 Low 1 3.60 我对如何在不包裹母亲的数据的情况下计算繁殖的概率感到困惑。我最接近这个数字的是没有平滑线的逻辑回归图。这是我的代码:https://link.springer.com/article/10.1007/s00442-009-1522-7#Fig2 ggplot(x, aes(Final_Size, Embryo_Presence_vs._Absence)) + geom_jitter(height = 0.05) + stat_smooth(method="glm", se= FALSE, method.args = list(family=binomial))

即使我以某种方式完成了这个数字,它不会计算繁殖的概率。有人可以读到可以计算经验概率的软件包或功能吗?我发现了许多有关概率密度图的帖子,我迷失在接下来的位置上。
	

您要模拟的数字显示了预期值,我们可以通过将模型与glm()Poor logistic regression plot拟合来计算出来。

snails <- glm( # use * instead of + to allow odds ratios to differ b/t treatments Embryo_Presence_vs._Absence ~ P_Treatment * Final_Size, family = binomial(link = "logit"), data = x )

当构建图时,我们可以使用
ggplot2 probability logistic-regression
1个回答
0
投票
而不是观察到的值(

fitted.values(snails)

)将
Embryo_Presence_vs._Absence

美学映射到预期值。例如:
ggplot(x, aes(Final_Size, fitted.values(snails),)) + 
  geom_point(aes(shape = P_Treatment)) +
  scale_shape_manual(values = c(1, 16)) + theme_classic()

    

最新问题
© www.soinside.com 2019 - 2025. All rights reserved.