Sequences detection: cicada_tools.sequences¶
from cicada_analysis.cicada_tools.sequences.<module> import <function>
Sequence detection from calcium imaging data (Villette et al. 2015). Not re-exported at package level: import from the submodule:
from cicada_analysis.cicada_tools.sequences.malvache.utils_malvache_sequences import step1_pca_off_memory
sequences.malvache.utils_malvache_sequences (PCA-based sequence detection)¶
Used by the cicada_seq_malvache_analysis plugin:
-
step1_pca_off_memory¶
step1_pca_off_memory(data, n_pc_max) -> (data_centralized, pc_time_course)
PCA on the lag-1 temporal covariance matrix, matrix-free (ARPACK): avoids forming the full n_times × n_times covariance matrix.
| Parameter | Type | Description |
|---|---|---|
data |
np.ndarray | Input data |
n_pc_max |
int | Maximum number of PCs |
Returns: (data_centralized, pc_time_course)
-
covnorm_vectorized¶
covnorm_vectorized(A, b, lags) -> np.ndarray
Normalized cross-covariance between each row of A and b, for lags -lags to +lags.
| Parameter | Type | Description |
|---|---|---|
A |
np.ndarray | Signal matrix |
b |
np.ndarray | Target signal |
lags |
int | Number of lags |
Returns: np.ndarray — Vectorized cross-covariance
-
centralize_data¶
centralize_data(data) -> np.ndarray
Double-centers a (cells × times) matrix.
| Parameter | Type | Description |
|---|---|---|
data |
np.ndarray | Input matrix |
Returns: np.ndarray — Centered matrix
-
convert_position_to_distance¶
convert_position_to_distance(positions, loop_length) -> (total_traveled_distance, n_laps)
Converts a circular-track position trace into cumulative distance traveled.
| Parameter | Type | Description |
|---|---|---|
positions |
np.ndarray | Position trace |
loop_length |
float | Loop length |
Returns: (total_traveled_distance, n_laps)
-
normalize_array_0_255¶
normalize_array_0_255(img_array) -> np.ndarray
Min-max rescaling to 0-255, cast to uint8 (truncated, not rounded), e.g. to save a matrix as an image. A constant array is only cast to uint8, not rescaled.
| Parameter | Type | Description |
|---|---|---|
img_array |
np.ndarray | Values to rescale |
Returns: np.ndarray — uint8, same shape