Details
Description
DistributedLanczosSolver (line 99) claims to persist eigenVectors.numRows() vectors.
log.info("Persisting " + eigenVectors.numRows() + " eigenVectors and eigenValues to: " + outputPath);
However, a few lines later (line 106) we have
for(int i=0; i<eigenVectors.numRows() - 1; i++) { ... }
which only persists eigenVectors.numRows()-1 vectors.
Seems like the most significant eigenvector (i.e. the one with the largest eigenvalue) is omitted... off by one bug?
Also, I think it would be better if the eigenvectors are persisted in reverse order, meaning the most significant vector is marked "0", the 2nd most significant is marked "1", etc.
This, for two reasons:
1) When performing another PCA on the same corpus (say, with more principal componenets), corresponding eigenvalues can be easily matched and compared.
2) Makes it easier to discard the least significant principal components, which for Lanczos decomposition are usually garbage.
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Field | Original Value | New Value |
---|---|---|
Description |
DistributedLanczosSolver (line 99) claims to persist eigenVectors.numRows() vectors.
{code} log.info("Persisting " + eigenVectors.numRows() + " eigenVectors and eigenValues to: " + outputPath); {code} However, a few lines later (line 106) we have {code} for(int i=0; i<eigenVectors.numRows() - 1; i++) { ... } {code} which only persists eigenVectors.numRows()-1 vectors. Seems like the most significant eigenvector (i.e. the one with the largest eigenvalue) is omitted... off by one bug? Also, I think it would be better if the eigenvectors are persisted in *reverse* order, meaning the most significant vector is marked "0", the 2nd most significant is marked "1", etc. This, for two reasons: 1) When performing another PCA on the same corpus (say, with more principal componenets), corresponding eigenvalues can be easily matched and compared. 2) Makes it easier to discard the least significant principal components, which for Lanczos decomposition are usually garabage. |
DistributedLanczosSolver (line 99) claims to persist eigenVectors.numRows() vectors.
{code} log.info("Persisting " + eigenVectors.numRows() + " eigenVectors and eigenValues to: " + outputPath); {code} However, a few lines later (line 106) we have {code} for(int i=0; i<eigenVectors.numRows() - 1; i++) { ... } {code} which only persists eigenVectors.numRows()-1 vectors. Seems like the most significant eigenvector (i.e. the one with the largest eigenvalue) is omitted... off by one bug? Also, I think it would be better if the eigenvectors are persisted in *reverse* order, meaning the most significant vector is marked "0", the 2nd most significant is marked "1", etc. This, for two reasons: 1) When performing another PCA on the same corpus (say, with more principal componenets), corresponding eigenvalues can be easily matched and compared. 2) Makes it easier to discard the least significant principal components, which for Lanczos decomposition are usually garbage. |
Fix Version/s | 0.4 [ 12314396 ] |
Attachment | MAHOUT-369.patch [ 12441827 ] |
Assignee | Jake Mannix [ jake.mannix ] |
Fix Version/s | 0.5 [ 12315255 ] | |
Fix Version/s | 0.4 [ 12314396 ] |
Attachment | MAHOUT-369.diff [ 12475343 ] |
Status | Open [ 1 ] | Patch Available [ 10002 ] |
Status | Patch Available [ 10002 ] | Resolved [ 5 ] |
Resolution | Fixed [ 1 ] |
Status | Resolved [ 5 ] | Closed [ 6 ] |