Michael-Geis
reorganized
7cc8002
import pandas as pd
from sklearn.pipeline import Pipeline
from storage import Fetch
from cleaning import TextCleaner
from embedding import Embedder
from search import Search
def get_recs(id_list, save_recs=False):
path_to_library = "./data/libraries/APSP_50_allenai-specter"
path_to_save_recs = "./output/"
## Create pipeline
model = Pipeline(
[
("fetch", Fetch()),
("clean", TextCleaner()),
("embed", Embedder(model_name="allenai-specter")),
("search", Search(path_to_library=path_to_library)),
]
)
recommendation_df = model.transform(id_list)
if save_recs:
recommendation_df.to_feather(path_to_save_recs)
return recommendation_df