Deduplicate 50k rows historical persons

Linking a dataset of real historical persons¶

In this example, we deduplicate a more realistic dataset. The data is based on historical persons scraped from wikidata. Duplicate records are introduced with a variety of errors introduced.

from splink.duckdb.duckdb_linker import DuckDBLinker
import altair as alt
alt.renderers.enable('mimetype')

import pandas as pd 
pd.options.display.max_rows = 1000
df = pd.read_parquet("./data/historical_figures_with_errors_50k.parquet")

# Simple settings dictionary will be used for exploratory analysis
settings = {
    "link_type": "dedupe_only",
    "blocking_rules_to_generate_predictions": [
        "l.first_name = r.first_name and l.surname = r.surname",
        "l.surname = r.surname and l.dob = r.dob",
        "l.first_name = r.first_name and l.dob = r.dob",
        "l.postcode_fake = r.postcode_fake and l.first_name = r.first_name",
    ],
}
linker = DuckDBLinker(df, settings)

linker.profile_columns(
    ["first_name", "postcode_fake", "substr(dob, 1,4)"], top_n=10, bottom_n=5
)

linker.cumulative_num_comparisons_from_blocking_rules_chart()

import splink.duckdb.duckdb_comparison_library as cl

settings = {
    "link_type": "dedupe_only",
    "blocking_rules_to_generate_predictions": [
        "l.first_name = r.first_name and l.surname = r.surname",
        "l.surname = r.surname and l.dob = r.dob",
        "l.first_name = r.first_name and l.dob = r.dob",
        "l.postcode_fake = r.postcode_fake and l.first_name = r.first_name",
    ],
    "comparisons": [
        cl.jaro_winkler_at_thresholds("first_name", [0.9, 0.7], term_frequency_adjustments=True),
        cl.jaro_winkler_at_thresholds("surname", [0.9, 0.7], term_frequency_adjustments=True),
        cl.levenshtein_at_thresholds("dob", [1,2], term_frequency_adjustments=True),
        cl.levenshtein_at_thresholds("postcode_fake", 2,term_frequency_adjustments=True),
        cl.exact_match("birth_place", term_frequency_adjustments=True),
        cl.exact_match("occupation",  term_frequency_adjustments=True),
    ],
    "retain_matching_columns": True,
    "retain_intermediate_calculation_columns": True,
    "max_iterations": 10,
    "em_convergence": 0.01
}

linker = DuckDBLinker(df, settings, connection=":temporary:")

linker.estimate_probability_two_random_records_match(
    [
        "l.first_name = r.first_name and l.surname = r.surname and l.dob = r.dob",
        "substr(l.first_name,1,2) = substr(r.first_name,1,2) and l.surname = r.surname and substr(l.postcode_fake,1,2) = substr(r.postcode_fake,1,2)",
        "l.dob = r.dob and l.postcode_fake = r.postcode_fake",
    ],
    recall=0.6,
)

Probability two random records match is estimated to be  0.000136.
This means that amongst all possible pairwise record comparisons, one in 7,362.31 are expected to match.  With 1,279,041,753 total possible comparisons, we expect a total of around 173,728.33 matching pairs

linker.estimate_u_using_random_sampling(target_rows=5e6)

----- Estimating u probabilities using random sampling -----

Estimated u probabilities using random sampling

Your model is not yet fully trained. Missing estimates for:
    - first_name (no m values are trained).
    - surname (no m values are trained).
    - dob (no m values are trained).
    - postcode_fake (no m values are trained).
    - birth_place (no m values are trained).
    - occupation (no m values are trained).

blocking_rule = "l.first_name = r.first_name and l.surname = r.surname"
training_session_names = linker.estimate_parameters_using_expectation_maximisation(blocking_rule)


----- Starting EM training session -----

Estimating the m probabilities of the model by blocking on:
l.first_name = r.first_name and l.surname = r.surname

Parameter estimates will be made for the following comparison(s):
    - dob
    - postcode_fake
    - birth_place
    - occupation

Parameter estimates cannot be made for the following comparison(s) since they are used in the blocking rules: 
    - first_name
    - surname

Iteration 1: Largest change in params was -0.524 in probability_two_random_records_match
Iteration 2: Largest change in params was -0.0339 in probability_two_random_records_match
Iteration 3: Largest change in params was 0.0119 in the m_probability of birth_place, level `Exact match`
Iteration 4: Largest change in params was 0.00374 in the m_probability of birth_place, level `Exact match`

EM converged after 4 iterations

Your model is not yet fully trained. Missing estimates for:
    - first_name (no m values are trained).
    - surname (no m values are trained).

blocking_rule = "l.dob = r.dob"
training_session_dob = linker.estimate_parameters_using_expectation_maximisation(blocking_rule)


----- Starting EM training session -----

Estimating the m probabilities of the model by blocking on:
l.dob = r.dob

Parameter estimates will be made for the following comparison(s):
    - first_name
    - surname
    - postcode_fake
    - birth_place
    - occupation

Parameter estimates cannot be made for the following comparison(s) since they are used in the blocking rules: 
    - dob

Iteration 1: Largest change in params was -0.362 in the m_probability of first_name, level `Exact match`
Iteration 2: Largest change in params was -0.0302 in the m_probability of first_name, level `Exact match`
Iteration 3: Largest change in params was 0.00453 in the m_probability of first_name, level `All other comparisons`

EM converged after 3 iterations

Your model is fully trained. All comparisons have at least one estimate for their m and u values

The final match weights can be viewed in the match weights chart:

linker.match_weights_chart()

linker.unlinkables_chart()

df_predict = linker.predict()
df_e = df_predict.as_pandas_dataframe(limit=5)
df_e

	match_weight	match_probability	unique_id_l	unique_id_r	first_name_l	first_name_r	gamma_first_name	tf_first_name_l	tf_first_name_r	bf_first_name	...	bf_birth_place	bf_tf_adj_birth_place	occupation_l	occupation_r	gamma_occupation	tf_occupation_l	tf_occupation_r	bf_occupation	bf_tf_adj_occupation
0	18.131419	0.999997	Q2296770-1	Q2296770-14	thomas	thomas	3	0.028667	0.028667	44.380695	...	1.000000	1.000000	politician	politician	1	0.088932	0.088932	22.786409	0.443341
1	40.311237	1.000000	Q1443188-1	Q1443188-3	frank	frank	3	0.006335	0.006335	44.380695	...	0.154553	1.000000	liturgist	liturgist	1	0.000237	0.000237	22.786409	166.178966
2	40.311237	1.000000	Q1443188-2	Q1443188-3	frank	frank	3	0.006335	0.006335	44.380695	...	0.154553	1.000000	liturgist	liturgist	1	0.000237	0.000237	22.786409	166.178966
3	41.258395	1.000000	Q90404618-1	Q90404618-3	emlie	emlie	3	0.000099	0.000099	44.380695	...	178.565257	4.701184	playwright	playwright	1	0.002728	0.002728	22.786409	14.450345
4	41.258395	1.000000	Q90404618-2	Q90404618-3	emlie	emlie	3	0.000099	0.000099	44.380695	...	178.565257	4.701184	playwright	playwright	1	0.002728	0.002728	22.786409	14.450345

5 rows × 47 columns

You can also view rows in this dataset as a waterfall chart as follows:

from splink.charts import waterfall_chart
records_to_plot = df_e.to_dict(orient="records")
linker.waterfall_chart(records_to_plot, filter_nulls=False)

clusters = linker.cluster_pairwise_predictions_at_threshold(df_predict, threshold_match_probability=0.95)

Completed iteration 1, root rows count 638
Completed iteration 2, root rows count 122
Completed iteration 3, root rows count 35
Completed iteration 4, root rows count 6
Completed iteration 5, root rows count 0

linker.cluster_studio_dashboard(df_predict, clusters, "50k_cluster.html", sampling_method='by_cluster_size', overwrite=True)

from IPython.display import IFrame

IFrame(
    src="./50k_cluster.html", width="100%", height=1200
)

linker.roc_chart_from_labels_column("cluster",match_weight_round_to_nearest=0.02)

records = linker.prediction_errors_from_labels_column(
    "cluster",
    threshold=0.999,
    include_false_negatives=False,
    include_false_positives=True,
).as_record_dict()
linker.waterfall_chart(records)

# Some of the false negatives will be because they weren't detected by the blocking rules
records = linker.prediction_errors_from_labels_column(
    "cluster",
    threshold=0.5,
    include_false_negatives=True,
    include_false_positives=False,
).as_record_dict(limit=50)

linker.waterfall_chart(records)