--- title: Diginetica dataset keywords: fastai sidebar: home_sidebar summary: "Diginetica dataset." description: "Diginetica dataset." nb_path: "nbs/datasets/diginetica.ipynb" ---
{% raw %}
{% endraw %} {% raw %}
{% endraw %} {% raw %}

class DigineticaDataset[source]

DigineticaDataset(root, column_names={'SESSION_ID': 'sessionId', 'ITEM_ID': 'itemId', 'TIMEFRAME': 'timeframe', 'EVENT_DATE': 'eventdate'}) :: SessionDatasetv2

Dataset base class

{% endraw %} {% raw %}
{% endraw %} {% raw %}
ds = DigineticaDataset(root='/content/digidata')
Downloading https://github.com/RecoHut-Datasets/diginetica/raw/main/train-item-views.csv
Processing...
-- Reading data
Splitting date 1464134400.0
186670
15979
[('4737', 1451606400.0), ('4741', 1451606400.0), ('4742', 1451606400.0)]
[('289', 1464220800.0), ('290', 1464220800.0), ('302', 1464220800.0)]
-- Splitting train set and test set
43098
719470
60858
[[1], [3, 4], [3]] [1451606400.0, 1451606400.0, 1451606400.0] [2, 5, 4]
[[21553, 20071, 8762, 21566, 6381], [21553, 20071, 8762, 21566], [21553, 20071, 8762]] [1464220800.0, 1464220800.0, 1464220800.0] [21566, 6381, 21566]
avg length:  4.850942344040704
Done!
{% endraw %} {% raw %}

class DigineticaDatasetv2[source]

DigineticaDatasetv2(root, shuffle=False, n_node=43097, is_train=True) :: SessionGraphDataset

References

1. COTREC session-based recommender model training. https://t.ly/cXTH.
{% endraw %} {% raw %}
{% endraw %} {% raw %}
root = '/content/diginetica'

train_data = DigineticaDataset(root=root, shuffle=True, is_train=True)
test_data = DigineticaDataset(root=root, shuffle=False, is_train=False)
{% endraw %}