Use rules in textual form

In this tutorial, we will load a set of regression rules in textual form and evaluate them

Load and prepare dataset

We begin by loading the boston-housing dataset into a DataFrame.

import pandas as pd
BOSTON_HOUSING_PATH = (
    'https://raw.githubusercontent.com/ruleminer/decision-rules/'
    'refs/heads/docs/docs-src/source/tutorials/resources/boston-housing.csv'
)
boston_housing_df = pd.read_csv(BOSTON_HOUSING_PATH)
display(boston_housing_df)
print('Columns: ', boston_housing_df.columns.values)
X = boston_housing_df.drop("MEDV", axis=1)
y = boston_housing_df["MEDV"]

	index	CRIM	ZN	INDUS	CHAS	NOX	RM	AGE	DIS	RAD	TAX	PTRATIO	B	LSTAT	MEDV
0	0	0.00632	18	2.31	0	0.538	6.575	65.2	4.0900	1	296	15	396.90	4.98	24.0
1	1	0.02731	0	7.07	0	0.469	6.421	78.9	4.9671	2	242	17	396.90	9.14	21.6
2	2	0.02729	0	7.07	0	0.469	7.185	61.1	4.9671	2	242	17	392.83	4.03	34.7
3	3	0.03237	0	2.18	0	0.458	6.998	45.8	6.0622	3	222	18	394.63	2.94	33.4
4	4	0.06905	0	2.18	0	0.458	7.147	54.2	6.0622	3	222	18	396.90	5.33	36.2
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
501	501	0.06263	0	11.93	0	0.573	6.593	69.1	2.4786	1	273	21	391.99	9.67	22.4
502	502	0.04527	0	11.93	0	0.573	6.120	76.7	2.2875	1	273	21	396.90	9.08	20.6
503	503	0.06076	0	11.93	0	0.573	6.976	91.0	2.1675	1	273	21	396.90	5.64	23.9
504	504	0.10959	0	11.93	0	0.573	6.794	89.3	2.3889	1	273	21	393.45	6.48	22.0
505	505	0.04741	0	11.93	0	0.573	6.030	80.8	2.5050	1	273	21	396.90	7.88	11.9

506 rows × 15 columns

Columns:  ['index' 'CRIM' 'ZN' 'INDUS' 'CHAS' 'NOX' 'RM' 'AGE' 'DIS' 'RAD' 'TAX'
 'PTRATIO' 'B' 'LSTAT' 'MEDV']

Load the ruleset in textual form

Now we need to load the ruleset provided in a text file

import urllib

FILE_PATH: str = (
    'https://raw.githubusercontent.com/ruleminer/decision-rules/'
    'refs/heads/docs/docs-src/source/tutorials/resources/regression/text_ruleset.txt'
)

with urllib.request.urlopen(FILE_PATH) as response:
    text_rules_model = response.read().decode('utf-8').splitlines()

text_rules_model

['IF AGE >= 80.05 AND RM < 7.20 AND LSTAT >= 14.74 AND CRIM >= 1.06',
 'IF LSTAT >= 14.43 AND AGE >= 77.95 AND CRIM >= 0.24',
 'IF TAX >= 300.00 AND CRIM < 15.72 AND RM >= 5.06 AND LSTAT < 32.00 AND LSTAT >= 14.73',
 'IF RM < 6.45 AND RM >= 5.75 AND AGE < 91.05 AND LSTAT < 14.16',
 'IF RM < 6.59 AND B >= 198.44 AND LSTAT < 16.12 AND RM >= 5.64 AND DIS >= 1.15']

Convert the textual ruleset to a decision-rules model

Now that the rules are loaded, we convert them into a decision-rules model using the TextRulesetFactory from decision-rules library. This conversion enables us to evaluate and modify the ruleset programmatically.

from decision_rules.ruleset_factories._factories.regression import TextRuleSetFactory

factory = TextRuleSetFactory()
ruleset = factory.make(text_rules_model, X, y)

After conversion in the decision-rules library, we can easilythe display the model

for rule in ruleset.rules:
    print(rule)

IF AGE >= 80.05 AND RM < 7.20 AND LSTAT >= 14.74 AND CRIM >= 1.06 THEN MEDV = {13.13} [9.41, 16.84] (p=72, n=31, P=105, N=401)
IF LSTAT >= 14.43 AND AGE >= 77.95 AND CRIM >= 0.24 THEN MEDV = {14.08} [10.07, 18.10] (p=100, n=36, P=125, N=381)
IF TAX >= 300.00 AND CRIM < 15.72 AND RM >= 5.06 AND LSTAT < 32.00 AND LSTAT >= 14.73 THEN MEDV = {15.23} [11.46, 19.00] (p=89, n=32, P=139, N=367)
IF RM < 6.45 AND RM >= 5.75 AND AGE < 91.05 AND LSTAT < 14.16 THEN MEDV = {22.05} [18.49, 25.62] (p=129, n=11, P=227, N=279)
IF RM < 6.59 AND B >= 198.44 AND LSTAT < 16.12 AND RM >= 5.64 AND DIS >= 1.15 THEN MEDV = {21.75} [18.05, 25.45] (p=184, n=33, P=236, N=270)

Analyze the ruleset statistics

We can compute various metrics for the ruleset. This step involves retrieving statistical information about the rules.

We start by calculating and displaying the general characteristics of the ruleset

ruleset_stats = ruleset.calculate_ruleset_stats(X, y)
print(ruleset_stats)

{'rules_count': 5, 'avg_conditions_count': 4.2, 'avg_precision': 0.79, 'avg_coverage': 0.69, 'total_conditions_count': 21}

Now let’s calculate metrics for each rule. To make the output more readable and easier to interpret, we will organize the metrics into a DataFrame

rule_metrics = ruleset.calculate_rules_metrics(X, y)
rule_metrics_df = pd.DataFrame([
    {
        'Rule': f"r{i+1}",
        'p': metrics['p'],
        'n': metrics['n'],
        'P': metrics['P'],
        'N': metrics['N'],
        'unique_in_pos': metrics['unique_in_pos'],
        'unique_in_neg': metrics['unique_in_neg'],
        'p_unique': metrics['p_unique'],
        'n_unique': metrics['n_unique'],
        'all_unique': metrics['all_unique'],
        'support':round(metrics['support'],3),
        'conditions_count': metrics['conditions_count'],
        'y_covered_avg': round(metrics['y_covered_avg'],3),
        'y_covered_median': round(metrics['y_covered_median'],3),
        'y_covered_min': metrics['y_covered_min'],
        'y_covered_max': metrics['y_covered_max'],
        'mae': round(metrics['mae'],3),
        'rmse': round(metrics['rmse'],3),
        'mape': round(metrics['mape'],3),
        'p-value': round(metrics['p-value'],3)

    }
    for i, (_, metrics) in enumerate(rule_metrics.items())
])
display(rule_metrics_df)

	Rule	p	n	P	N	unique_in_pos	unique_in_neg	p_unique	n_unique	all_unique	support	conditions_count	y_covered_avg	y_covered_median	y_covered_min	y_covered_max	mae	rmse	mape	p-value
0	r1	72	31	105	401	3	6	0	0	0	0.204	4	13.128	13.40	5.0	27.5	10.096	13.148	0.398	0.0
1	r2	100	36	125	381	10	7	3	3	6	0.269	3	14.081	13.95	5.0	30.7	9.397	12.484	0.374	0.0
2	r3	89	32	139	367	13	18	10	11	21	0.239	5	15.233	14.90	6.3	27.5	8.650	11.735	0.351	0.0
3	r4	129	11	227	279	1	4	1	2	3	0.277	4	22.053	21.70	11.9	50.0	6.577	9.201	0.352	0.0
4	r5	184	33	236	270	53	25	43	15	58	0.429	5	21.747	21.40	11.9	50.0	6.549	9.222	0.346	0.0

We can also calculate statistics like condition importances

from decision_rules.measures import c2
condition_importances = ruleset.calculate_condition_importances(X, y, measure=c2)
condition_importances

[{'condition': 'LSTAT >= 14.43',
  'attributes': ['LSTAT'],
  'importance': 0.32767459373187646},
 {'condition': 'LSTAT >= 14.74',
  'attributes': ['LSTAT'],
  'importance': 0.16689142510435745},
 {'condition': 'LSTAT >= 14.73',
  'attributes': ['LSTAT'],
  'importance': 0.15660227492592682},
 {'condition': 'RM < 6.45',
  'attributes': ['RM'],
  'importance': 0.1464841812280583},
 {'condition': 'LSTAT < 14.16',
  'attributes': ['LSTAT'],
  'importance': 0.14473028867871895},
 {'condition': 'RM < 6.59',
  'attributes': ['RM'],
  'importance': 0.1174108026185524},
 {'condition': 'LSTAT < 16.12',
  'attributes': ['LSTAT'],
  'importance': 0.1090749271904711},
 {'condition': 'AGE < 91.05',
  'attributes': ['AGE'],
  'importance': 0.08332534928637518},
 {'condition': 'CRIM >= 1.06',
  'attributes': ['CRIM'],
  'importance': 0.07594939786770724},
 {'condition': 'AGE >= 77.95',
  'attributes': ['AGE'],
  'importance': 0.051180099750744834},
 {'condition': 'RM < 7.20',
  'attributes': ['RM'],
  'importance': 0.04416566101660634},
 {'condition': 'AGE >= 80.05',
  'attributes': ['AGE'],
  'importance': 0.03676886132555876},
 {'condition': 'CRIM < 15.72',
  'attributes': ['CRIM'],
  'importance': 0.030665695828604723},
 {'condition': 'RM >= 5.75',
  'attributes': ['RM'],
  'importance': 0.030479048373156292},
 {'condition': 'B >= 198.44',
  'attributes': ['B'],
  'importance': 0.028619349699875198},
 {'condition': 'RM >= 5.64',
  'attributes': ['RM'],
  'importance': 0.02736509389238543},
 {'condition': 'TAX >= 300.00',
  'attributes': ['TAX'],
  'importance': 0.02714594443889693},
 {'condition': 'CRIM >= 0.24',
  'attributes': ['CRIM'],
  'importance': 0.020564473495815715},
 {'condition': 'RM >= 5.06',
  'attributes': ['RM'],
  'importance': 0.01044814267795384},
 {'condition': 'LSTAT < 32.00',
  'attributes': ['LSTAT'],
  'importance': 0.003366620183006248},
 {'condition': 'DIS >= 1.15',
  'attributes': ['DIS'],
  'importance': 0.003149404172213438}]

Modify the ruleset

The decision-rule model can be easily edited. For example, we will create a new rule stating “IF RM < 6.95 AND TAX >= 219.00 AND LSTAT < 14.17 THEN MEDV = 23.35” and then add it to the ruleset.

from decision_rules.regression.rule import RegressionConclusion
from decision_rules.regression.rule import RegressionRule
from decision_rules.conditions import ElementaryCondition, CompoundCondition

rule = RegressionRule(
    premise=CompoundCondition(
        subconditions=[
            # Condition:  RM < 6.95
            ElementaryCondition(
                column_index=X.columns.get_loc('RM'),
                left=float('-inf'),
                right=6.95,
                left_closed=False,
                right_closed=False
            ),
            # Condition: TAX >= 219.00
            ElementaryCondition(
                column_index=X.columns.get_loc('TAX'),
                left=219.00,
                right=float('inf'),
                left_closed=True,
                right_closed=False
            ),
            # Condition: LSTAT < 14.17
            ElementaryCondition(
                column_index=X.columns.get_loc('LSTAT'),
                left=float('-inf'),
                right=14.17,
                left_closed=False,
                right_closed=False
            ),
        ]
    ),
    conclusion=RegressionConclusion(
        value=23.35,
        column_name='MEDV',
        low = 20.0,
        high = 25.0
    ),
    column_names=X.columns,
)
print(rule)

IF RM < 6.95 AND TAX >= 219.00 AND LSTAT < 14.17 THEN MEDV = {23.35} [20.00, 25.00]

rule.coverage = rule.calculate_coverage(X.to_numpy(), y.to_numpy())
print(rule.coverage)

(p=203, n=42, P=255, N=251)

ruleset.rules.append(rule)

print("Updated Ruleset:")
for rule in ruleset.rules:
    print(rule)

Updated Ruleset:
IF AGE >= 80.05 AND RM < 7.20 AND LSTAT >= 14.74 AND CRIM >= 1.06 THEN MEDV = {13.13} [9.41, 16.84] (p=72, n=31, P=105, N=401)
IF LSTAT >= 14.43 AND AGE >= 77.95 AND CRIM >= 0.24 THEN MEDV = {14.08} [10.07, 18.10] (p=100, n=36, P=125, N=381)
IF TAX >= 300.00 AND CRIM < 15.72 AND RM >= 5.06 AND LSTAT < 32.00 AND LSTAT >= 14.73 THEN MEDV = {15.23} [11.46, 19.00] (p=89, n=32, P=139, N=367)
IF RM < 6.45 AND RM >= 5.75 AND AGE < 91.05 AND LSTAT < 14.16 THEN MEDV = {22.05} [18.49, 25.62] (p=129, n=11, P=227, N=279)
IF RM < 6.59 AND B >= 198.44 AND LSTAT < 16.12 AND RM >= 5.64 AND DIS >= 1.15 THEN MEDV = {21.75} [18.05, 25.45] (p=184, n=33, P=236, N=270)
IF RM < 6.95 AND TAX >= 219.00 AND LSTAT < 14.17 THEN MEDV = {23.28} [18.17, 28.39] (p=203, n=42, P=255, N=251)

Now let’s remove from the rule “IF LSTAT >= 14.43 AND AGE >= 77.95 AND CRIM >= 0.24 THEN MEDV = {14.08} [10.07, 18.10]” the condition “AGE >= 77.95”

condition_to_remove = ruleset.rules[1].premise.subconditions[1]
ruleset.rules[1].premise.subconditions.remove(condition_to_remove)
ruleset.rules[1].coverage = ruleset.rules[1].calculate_coverage(X.to_numpy(), y.to_numpy())

print("Updated Ruleset:")
for rule in ruleset.rules:
    print(rule)

Updated Ruleset:
IF AGE >= 80.05 AND RM < 7.20 AND LSTAT >= 14.74 AND CRIM >= 1.06 THEN MEDV = {13.13} [9.41, 16.84] (p=72, n=31, P=105, N=401)
IF LSTAT >= 14.43 AND CRIM >= 0.24 THEN MEDV = {14.22} [10.01, 18.44] (p=105, n=41, P=134, N=372)
IF TAX >= 300.00 AND CRIM < 15.72 AND RM >= 5.06 AND LSTAT < 32.00 AND LSTAT >= 14.73 THEN MEDV = {15.23} [11.46, 19.00] (p=89, n=32, P=139, N=367)
IF RM < 6.45 AND RM >= 5.75 AND AGE < 91.05 AND LSTAT < 14.16 THEN MEDV = {22.05} [18.49, 25.62] (p=129, n=11, P=227, N=279)
IF RM < 6.59 AND B >= 198.44 AND LSTAT < 16.12 AND RM >= 5.64 AND DIS >= 1.15 THEN MEDV = {21.75} [18.05, 25.45] (p=184, n=33, P=236, N=270)
IF RM < 6.95 AND TAX >= 219.00 AND LSTAT < 14.17 THEN MEDV = {23.28} [18.17, 28.39] (p=203, n=42, P=255, N=251)

We can also modify the value of a condition. In the rule “IF RM < 6.45 AND RM >= 5.75 AND AGE < 91.05 AND LSTAT < 14.16 THEN MEDV = {22.05} [18.49, 25.62]” we will update the condition “AGE < 91.05” to “AGE <= 71.5”

ruleset.rules[3].premise.subconditions[2].right = 71.5
ruleset.rules[3].premise.subconditions[2].right_closed = True
ruleset.rules[3].coverage = ruleset.rules[3].calculate_coverage(X.to_numpy(), y.to_numpy())

print("Updated Ruleset:")
for rule in ruleset.rules:
    print(rule)

Updated Ruleset:
IF AGE >= 80.05 AND RM < 7.20 AND LSTAT >= 14.74 AND CRIM >= 1.06 THEN MEDV = {13.13} [9.41, 16.84] (p=72, n=31, P=105, N=401)
IF LSTAT >= 14.43 AND CRIM >= 0.24 THEN MEDV = {14.22} [10.01, 18.44] (p=105, n=41, P=134, N=372)
IF TAX >= 300.00 AND CRIM < 15.72 AND RM >= 5.06 AND LSTAT < 32.00 AND LSTAT >= 14.73 THEN MEDV = {15.23} [11.46, 19.00] (p=89, n=32, P=139, N=367)
IF RM < 6.45 AND RM >= 5.75 AND AGE <= 71.50 AND LSTAT < 14.16 THEN MEDV = {22.28} [19.69, 24.87] (p=85, n=22, P=173, N=333)
IF RM < 6.59 AND B >= 198.44 AND LSTAT < 16.12 AND RM >= 5.64 AND DIS >= 1.15 THEN MEDV = {21.75} [18.05, 25.45] (p=184, n=33, P=236, N=270)
IF RM < 6.95 AND TAX >= 219.00 AND LSTAT < 14.17 THEN MEDV = {23.28} [18.17, 28.39] (p=203, n=42, P=255, N=251)