Loading a Model

Load Data & Model

Assume you have trained a model and stored it under demonstration_model. We can now load the model and use it to make predictions using left and right DataFrames. In the first step, both the model and the data are loaded into memory.

import pandas as pd

from neer_match_utilities.model import Model
from neer_match_utilities.prepare import Prepare, similarity_map_to_dict
from pathlib import Path

# Load files

left = pd.read_csv('left.csv')
right = pd.read_csv('right.csv')

# Load model

loaded_model = Model.load(
    'demonstration_model'
)

Harmonize Format

Next, we must ensure that the formatting logic remains consistent with that applied before training. Note that it is not necessary to redefine the similarity map, as it was stored and is loaded along with the model.

prepare = Prepare(
    similarity_map=similarity_map_to_dict(
        loaded_model.similarity_map
    ), 
    df_left=left, 
    df_right=right, 
    id_left='company_id', 
    id_right='company_id'
)

left, right = prepare.format(
    fill_numeric_na=False,
    to_numeric=['found_year'],
    fill_string_na=True, 
    capitalize=True
)

Make Suggestions

Now we can make suggestions:

# Make suggestions for the first observation in left

suggestions = loaded_model.suggest(
    left[:1], 
    right, 
    count=10, 
    verbose=0
)

suggestions

	left	right	prediction
0	0	0	0.473703
411	0	411	0.381956
675	0	675	0.362005
256	0	256	0.347396
497	0	497	0.345034
439	0	439	0.341827
132	0	132	0.323066
529	0	529	0.322083
181	0	181	0.319886
633	0	633	0.319725

Based on this output, we can assess whether the suggestion is correct.

left.iloc[0]

company_id                                             1e87fc75b4
company_name    GLÜCKAUF-, ACTIEN-GESELLSCHAFT FÜR BRAUNKOHLEN...
city                                                    LICHTENAU
industry                     BERGWERKE, HÜTTEN- UND SALINENWESEN.
purpose         ABBAU VON BRAUNKOHLENLAGERN U. BRIKETTFABRIKAT...
bs_text         GRUNDST CKE M GRUBENWERT M SCHACHTANLAGEN M GE...
found_year                                                 1871.0
Name: 0, dtype: object

right.iloc[
    suggestions.loc[0, 'right']
]

company_id                                             0008e07878
company_name    „GLÜCKAUF-', ACT.-GES. FÜR BRAUNKOHLEN-VERWERT...
city                                                             
industry                                                NACHTRAG.
purpose         ABBAU VON BRAUNKOHLENLAGERN U. BRIKETTFABRIKAT...
bs_text         GRUNDST CKE M GRUBENWERT M SCHACHTANLAGEN M GE...
found_year                                                 1871.0
Name: 0, dtype: object