sequenzo 0.1.20__cp39-cp39-win_amd64.whl → 0.1.22__cp39-cp39-win_amd64.whl

sequenzo/dissimilarity_measures/utils/seqconc.pyx

@@ -1,26 +0,0 @@
-import numpy as np
-cimport numpy as cnp
-from libc.stdint cimport int32_t
-from libc.string cimport strcmp
-
-cdef str sconc_np(cnp.ndarray[int32_t, ndim=1] seqdata, str sep):
-    cdef int i, size = seqdata.shape[0]
-    cdef list valid_values = []
-    cdef bytes result = b""
-
-    for i in range(size):
-        if seqdata[i] >= 0:
-            valid_values.append(str(seqdata[i]))
-
-    if valid_values:
-        result = sep.join(valid_values).encode('utf-8')
-
-    return result.decode('utf-8')
-
-def seqconc(cnp.ndarray[int32_t, ndim=2] data, str sep="-"):
-    if data.ndim == 1:
-        return sconc_np(data, sep)
-    elif data.ndim == 2:
-        return np.array([sconc_np(row, sep) for row in data])
-    else:
-        raise ValueError("Only 1D and 2D arrays are supported.")

sequenzo/dissimilarity_measures/utils/seqdss.pyx

@@ -1,33 +0,0 @@
-import numpy as np
-cimport numpy as cnp
-from sequenzo.define_sequence_data import SequenceData
-from libc.stdint cimport int32_t
-
-def seqdss(seqdata):
-    if not isinstance(seqdata, SequenceData):
-        raise ValueError("[!] data is NOT a sequence object, see SequenceData to create one.")
-
-    cdef cnp.ndarray[int32_t, ndim=2] seqdatanum = seqdata.values.astype(np.int32, copy=False)
-    cdef int n = seqdatanum.shape[0], m = seqdatanum.shape[1]
-    cdef int i, j
-
-    cdef cnp.ndarray[int32_t, ndim=2] ffill = seqdatanum
-    cdef cnp.ndarray[char, ndim=2] boundaries = np.ones((n, 1), dtype=bool)  # Fixed here
-
-    for j in range(1, m):
-        mask = (ffill[:, j] < 0)
-        ffill[mask, j] = ffill[mask, j - 1]
-
-    boundaries = np.concatenate([boundaries, ffill[:, 1:] != ffill[:, :-1]], axis=1)
-
-    cdef list groups = [row[boundary & (row >= 0)] for row, boundary in zip(ffill, boundaries)]
-    cdef int max_groups = max(len(g) for g in groups)
-
-    # 使用int32能表示的缺失值标记，避免np.nan转换为int32时的警告
-    cdef cnp.ndarray[int32_t, ndim=2] result = np.full((n, max_groups), -999, dtype=np.int32)
-
-    for i in range(n):
-        g = groups[i]
-        result[i, :len(g)] = g
-
-    return result

sequenzo/dissimilarity_measures/utils/seqdur.pyx

@@ -1,34 +0,0 @@
-import numpy as np
-cimport numpy as cnp
-from sequenzo.define_sequence_data import SequenceData
-from libc.stdint cimport int32_t
-
-def seqdur(seqdata):
-    if not isinstance(seqdata, SequenceData):
-        raise ValueError("data is not a sequence object, see SequenceData to create one")
-
-    cdef cnp.ndarray[int32_t, ndim=2] seqdatanum = seqdata.values.copy().astype(np.int32, copy=False)
-
-    cdef int n = seqdatanum.shape[0], m = seqdatanum.shape[1]
-    cdef int i, j
-
-    cdef cnp.ndarray[int32_t, ndim=2] ffill = seqdatanum.copy()
-
-    for j in range(1, m):
-        ffill[:, j] = np.where(seqdatanum[:, j] < 0, ffill[:, j - 1], seqdatanum[:, j])
-
-    cdef cnp.ndarray[char, ndim=2] boundaries = np.concatenate(
-        [np.ones((n, 1), dtype=bool), ffill[:, 1:] != ffill[:, :-1]], axis=1
-    )
-
-    cdef cnp.ndarray[int32_t, ndim=2] group_ids = np.cumsum(boundaries, axis=1).astype(np.int32, copy=False)
-    cdef cnp.ndarray[char, ndim=2] valid = seqdatanum >= 0
-
-    cdef cnp.ndarray[int32_t, ndim=2] group_durations = np.zeros((n, m), dtype=np.int32)
-
-    for i in range(n):
-        counts = np.bincount(group_ids[i][valid[i]])
-        counts = counts[1:] if counts.size > 0 else np.array([], dtype=int)
-        group_durations[i, :len(counts)] = counts
-
-    return group_durations

sequenzo/dissimilarity_measures/utils/seqlength.pyx

@@ -1,19 +0,0 @@
-import pandas as pd
-import numpy as np
-cimport numpy as cnp
-from sequenzo.define_sequence_data import SequenceData
-from libc.stdint cimport int32_t
-
-def seqlength(seqdata):
-    if isinstance(seqdata, SequenceData):
-        seqdata = seqdata.seqdata.replace(np.nan, -99)
-
-    cdef cnp.ndarray[int32_t, ndim=2] seqarray_long
-
-    if isinstance(seqdata, pd.DataFrame):
-        seqarray_long = seqdata.to_numpy(dtype=np.int32)
-        return np.sum(seqarray_long > 0, axis=1)
-
-    else:
-        seqarray_long = seqdata
-        return np.sum(seqarray_long > 0, axis=1)