I have a data frame that has a large number of variables that differ over time in the form of an index for them. I want to retrieve the sets of highly variable variables.
data & lt; - data.frame (time_series = c (1,2,3), score 1 = c (0.5, 0.4, 0.6), score 2 = c (0.3, 0.2, 0.1), score 3 = c (0.1, 0.4, 0.5) ), Score 4 = C (0.5, 0.2, 0.4), score 5 = C (0.1, 0.1, 0.2))
Two functions that should give the same result,
< Pre>libraries (statistics) with a #autocorrelation function, a data frame on ACF_rate lag 0 & lt; - ACF (data [2: length (name (data)), plot = false, lag.max = 0) #Simple Pearson correlation function. Cor_results & lt; - cor (data [2: length (name (data)), method = "pearson"
results in a simple (X_results> 0.6)
, But it starts losing the names of variables.
I am trying to extract the set of threshold boundaries from the dataframe on a large scale. I hope that I am missing some simple built functions I
Edit: I realized that Spearman would be an absolutely wrong act for this, because It will rank only with the time limit as well.
You do something like that (Cor_results> 0.6) - (cor_results> 0.6) - (cor_results & gt; length: (name (data)), method = "useless" x <-> ; Arr.ind = TRUE) x # row call # score 1 1 1 # scores 5 5 1 # scores 2 2 2 # 3 3 3 # scores 5 5 3 # scores 4 4 scorers 1 score 5 5 score 3 3 5 score 5 5 5 5
To get the matrix of the cells of the cells that meet your needs, To make you more readable, you can
x []
No comments:
Post a Comment