get_target_additive.RdThis function samples target variable according to the logistic model with additive impact
get_target_additive(
kmer_dat,
weights = NULL,
zero_weight = NULL,
binary = TRUE
)output of generate_kmer_data
a vector of weights of motifs' impact on the outcome. The
length of weights should be the same as the number of motifs provided
during sequences generation (it is the motifs parameter in the
generate_kmer_data function). If weights parameter is
NULL, then weights will be sampled from the uniform distribution on
0-1 interval. The probability of success for target sampling will be
calculated based on the formula provided in details section. Default to
NULL.
a single value denoting the weight of no-motifs case. If
NULL, then we sample the weight from the uniform distribution on the
[-2, -1] interval. Default to NULL.
logical, indicating whether the produced target variable should be binary or continuous.
a binary vector of target variable sampled based on additive model.
This function assumes the following additive binomial model:
\(g(EY) = w_0 + w_1 X_{m_1} + w_2 X_{m_2} + \ldots + w_m X_{m_m}\)
where \(w_1, \ldots, w_m\) are weights related to motifs.
In the case when weights is NULL we calculate the probabilities
based on the formula
\( exp(1 + x_i)/(1 + exp(1 + x_i))\) where xi denotes the sum of weights of motifs occurring in ith sequence.
n_seq <- 20
sequence_length <- 20
alph <- letters[1:4]
motifs <- generate_motifs(alph, 4, 4, 4, 6)
results <- generate_kmer_data(n_seq, sequence_length, alph,
motifs, n_injections = 4)
get_target_additive(results)
#> [1] 0 1 0 1 1 1 1 1 1 1 0 0 0 0 0 0 1 1 0 1