>>> from sklearn.decomposition import PCA
>>> from sklearn.cluster import KMeans
>>> from techminer2.factor_analysis.tfidf import terms_by_cluster_frame
>>> terms_by_cluster_frame(
... #
... # PARAMS:
... field="author_keywords",
... #
... # TF PARAMS:
... is_binary=True,
... cooc_within=1,
... #
... # TF-IDF PARAMS:
... norm=None,
... use_idf=False,
... smooth_idf=False,
... sublinear_tf=False,
... #
... # TERM PARAMS:
... top_n=20,
... occ_range=(None, None),
... gc_range=(None, None),
... custom_terms=None,
... #
... # DESOMPOSITION PARAMS:
... decomposition_estimator = PCA(
... n_components=5,
... whiten=False,
... svd_solver="auto",
... tol=0.0,
... iterated_power="auto",
... n_oversamples=10,
... power_iteration_normalizer="auto",
... random_state=0,
... ),
... clustering_estimator_or_dict = KMeans(
... n_clusters=6,
... init="k-means++",
... n_init=10,
... max_iter=300,
... tol=0.0001,
... algorithm="elkan",
... random_state=0,
... ),
... #
... # DATABASE PARAMS:
... root_dir="example/",
... database="main",
... year_filter=(None, None),
... cited_by_filter=(None, None),
... ).head()
0 ... 5
0 FINANCIAL_INCLUSION 03:0590 ... FINTECH 31:5168
1 CROWDFUNDING 03:0335 ...
2 CYBER_SECURITY 02:0342 ...
3 CASE_STUDY 02:0340 ...
4 BLOCKCHAIN 02:0305 ...
[5 rows x 6 columns]