Closed hashknot closed 7 years ago
@hashknot Any reason why you added another dataset to the test suite? We have a few sample datasets, but I worry that the repo is already getting pretty large. Can you build this test without the huge DfR dump?
@hashknot Other than that one formatting issue, and the size of the test dataset, looks great! For the test dataset, I'd suggest just paring down what you have -- reduce to a single document, maybe reduce to only a few dozen words within that document, etc.
78952ba uses a minimal test WoS dataset instead of the DfR dataset, and fixes the formatting issue in the test file.
@hashknot Awesome! If you're happy with this (looks great to me), go ahead and create a PR from bug/TETHNE-145
into develop.
Correctly export corpus metadata having non-ascii values to *_meta.csv file.
Comment out failing LDA model test (TETHNE-147).