A GO-PCA run.
A GO-PCA “run” consists of metadata (e.g., GO-PCA version,
timestamp), configuration data, intermediate results, and the signatures
generated.
The run does not contain the input data itself (i.e., the original
expression matrix and and the list of gene sets). This is so that the
file size of the output file can be kept small.
The run does contain all other information necessary to reproduce the
results from the raw data, including the parameter settings used and hash
values for the input data. Furthermore, it contains some intermediate
results (e.g., the PCA loadings matrix and the PC scores) which allow for
some additional analyses that can help gain insight into the data and the
signatures generated.
| Parameters: |
- sig_matrix (
GOPCASignatureMatrix) – The GO-PCA signature matrix generated.
- gopca_version (str) – The GO-PCA version.
- timestamp (str) – The timestamp.
- exec_time (float) – The execution time (in seconds).
- expression_hash (str) – Hash value for the expression input data.
- config_hashes (list of str) – Hash values for the configuration input data.
- genes (Iterable of str) – The genes in the analysis.
- samples (Iterable of str) – The samples in the analysis.
- W (
numpy.ndarray (np.float64)) – The PCA loading matrix; shape = (len(genes) x # PCs).
There must be a 1-to-1 correspondence between genes and the rows
of W.
- Y (
numpy.ndarray (np.float64)) – The PC score matrix; shape = (len(samples) x # PCs).
There must be a 1-to-1 correspondence between samples and the
rows of Y.
|
|---|
-
classmethod
read_pickle(path)
Read a run from a pickle file.
| Parameters: | path (str) – The pickle file. |
|---|
| Returns: | |
|---|
| Return type: | None |
|---|
-
write_pickle(path)
Save the current object to a pickle file.
| Parameters: | path (str) – The output file. |
|---|
| Returns: | |
|---|
| Return type: | None |
|---|