我想检查特定数字变量对R中字符向量的影响。
A B C D E F_PATHOLOGIC_TUMOR_STAGE AJCC_TUMOR_PATHOLOGIC_PT GRADE PRIMARY_SITE
TCGA.2J.AAB4.01 -0.2527 -1.2093 0.7556 -0.7071 -0.1310 Stage IIB T2 G2 Other (please specify)
TCGA.2J.AAB6.01 1.2080 0.4815 -0.3242 -0.7088 4.0856 Stage IIA T3 G2 Body of Pancreas
TCGA.2J.AAB8.01 0.7529 0.7916 0.0596 -0.5287 1.1315 Stage IIB T3 G3 Head of Pancreas
TCGA.2J.AAB9.01 -0.3248 -0.7941 -0.5940 0.3325 -0.4385 Stage IIB T3 G1 Head of Pancreas
TCGA.2J.AABA.01 -0.4154 0.2808 0.7133 -0.7088 -0.6147 Stage IIB T3 G2 Head of Pancreas
TCGA.2J.AABE.01 0.6566 -0.1300 0.5751 -0.7088 -0.7381 Stage IIA T3 G2 Body of Pancreas
如您所见，列A到E包含数值。 F_PATHOLOGIC_TUMOR_STAGE包含值Stage I，Stage IA，Stage IB，Stage IIA，Stage IIB，Stage III和Stage IV。< / p>
我希望独立地在A上看到每个因素E对F_PATHOLOGIC_TUMOR_STAGE的影响。当我运行以下代码时
glm(AJCC_PATHOLOGIC_TUMOR_STAGE ~ A, data = merged.data.frame.Stage, family = "binomial")
我收到以下错误：
Error in eval(family$initialize) : y values must be 0 <= y <= 1
然后我想使用anova来获得结果。可能导致问题的原因是什么？如何避免？