Question

我想对数据框中的相邻行（1-2,3-4,5-6）进行t检验。我想测试相邻行是否明显大于0.如何在整个数据框中执行此操作？像这样的t检验：

t.test(x=rowname,y=NULL,alternative="one.sided", alternative = "greater")

例如：hsa-miR-99a-5p.dataTissue和hsa-miR-99a-5p.dataSerum都大于0吗？

                              HEP015         mm7        s26  TxHEP-014  TxHEP-015  TxHEP-018      vs29
hsa-miR-99a-5p.dataTissue  0.8234702  1.18956279  0.6145471 -1.1804234 -2.9679366 -1.2382820 -1.856565
hsa-miR-99a-5p.dataSerum  -4.7975142 -1.79686065 -0.6652281 -3.0752460 -3.3742772 -3.8129578 -1.099334
hsa-miR-93-5p.dataTissue  -0.2943401 -0.02306152  0.3097169 -2.0754544  0.3720528  2.1457004  2.003187
hsa-miR-93-5p.dataSerum    1.2905805 -2.07963898 -1.7277941 -1.9811846 -0.7549379 -0.9258338 -2.319956
hsa-miR-92b-3p.dataTissue -0.5874168 -1.55601781  0.8656526 -0.3436976 -2.8644071  1.2470288  4.652237
hsa-miR-92b-3p.dataSerum  -1.9323028 -1.53377562 -0.9441492 -2.7669216 -2.5563865 -1.1790849  2.637090

Answer 1

假设它位于名为data.frame的{{1}}中，一种方法如下：

df

看来，你的library(dplyr) library(broom) row_results <- apply(df, 1, function(x) glance(t.test(x, alternative = "greater"))) row_results <- rbind_all(row_results) row_results$name <- rownames(df) # row_results # Source: local data frame [6 x 7] # # estimate statistic p.value parameter conf.low conf.high name # (dbl) (dbl) (dbl) (dbl) (dbl) (dbl) (chr) # 1 -0.6593753 -1.1179897 0.8468370 6 -1.8054369 Inf hsa-miR-99a-5p.dataTissue # 2 -2.6602026 -4.6508774 0.9982498 6 -3.7716602 Inf hsa-miR-99a-5p.dataSerum # 3 0.3482573 0.6410169 0.2726011 6 -0.7074507 Inf hsa-miR-93-5p.dataTissue # 4 -1.2141093 -2.5674955 0.9787610 6 -2.1329943 Inf hsa-miR-93-5p.dataSerum # 5 0.2019113 0.2219719 0.4158491 6 -1.5656549 Inf hsa-miR-92b-3p.dataTissue # 6 -1.1822187 -1.7248956 0.9323470 6 -2.5140465 Inf hsa-miR-92b-3p.dataSerum实际上是两个变量，因此，我会将它们分解，如下所示：

rowname

这允许我们library(tidyr) row_results <- separate(row_results, name, c("name", "type"), sep = "\\.")新的group_by变量并创建一个新的指标变量name：

both_significant

您可能希望将逻辑更改为row_results %>% group_by(name) %>% mutate(both_significant = all(p.value > 0.8))，但我想在结果中添加一些p.value < 0.05示例：

TRUE

测试相邻行是否从0开始显着

1 个答案: