factor-analyzer

library0.5.1pypypi✓ verified 85d ago

A factor analysis (FA) implementation in Python, providing exploratory factor analysis using principal component or maximum likelihood extraction with varimax/oblimin rotations. Version 0.5.1 supports Kaiser-Meyer-Olkin (KMO) test, Bartlett's test of sphericity, and parallel analysis. Released irregularly; last update 2022.

analytics

pip install factor-analyzer

github ↗homepage ↗

INSTALL

IMPORT

SIG · FACTOR-ANALYZER

factor-analyzer

analyticspythonv0.5.1

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

FactorAnalyzer

✓ from factor_analyzer import FactorAnalyzer

✗ from factor-analyzer import FactorAnalyzer

hyphen vs underscore

calculate_kmo

✓ from factor_analyzer.factor_analyzer import calculate_kmo

✗ from factor_analyzer import calculate_kmo

calculate_kmo is not exposed at top-level

calculate_bartlett_sphericity

✓ from factor_analyzer.factor_analyzer import calculate_bartlett_sphericity

Load data, check suitability, then fit FA with 3 factors and varimax rotation.

import pandas as pd from factor_analyzer import FactorAnalyzer from factor_analyzer.factor_analyzer import calculate_kmo, calculate_bartlett_sphericity df = pd.read_csv('data.csv') # Drop rows with missing values df.dropna(inplace=True) # Bartlett's test chi2, p = calculate_bartlett_sphericity(df) print('Bartlett p-value:', p) # KMO kmo_all, kmo_total = calculate_kmo(df) print('KMO:', kmo_total) # Factor analysis with 3 factors, varimax rotation fa = FactorAnalyzer(rotation='varimax', n_factors=3) fa.fit(df) print('Loadings:\n', fa.loadings_)

Debug

Known issues

gotchaFactorAnalyzer.fit() silently drops rows with NaN. Always check for missing values first.

fix

Use df.dropna() before fitting.

affects: <0.5.0

gotchaRotation parameter must be lowercase: 'varimax', not 'Varimax'. Unknown rotation raises AttributeError.

fix

Use rotation='varimax', 'oblimin', or None.

affects: all

breakingIn v0.4.0, the method 'get_communalities()' was removed. Use fa.communalities_ attribute directly.

fix

Access fa.communalities_ after fit.

affects: >=0.4.0

gotchacalculate_kmo returns (kmo_per_variable, kmo_total). Many users forget the tuple unpacking.

fix

kmo_all, kmo_total = calculate_kmo(df)

affects: all

deprecatedThe module factor_analyzer.factor_analyzer is internal; top-level imports are preferred for FactorAnalyzer class.

fix

Use from factor_analyzer import FactorAnalyzer

affects: >=0.5.0

Errors

Common errors & fixes

AttributeError: module 'factor_analyzer' has no attribute 'FactorAnalyzer'

Installed 'factor-analyzer' package but trying to import from 'factor-analyzer' with hyphen.

fix

Install correct package: pip install factor-analyzer ; import from factor_analyzer (underscore)

ValueError: n_features must be >= n_factors

Number of factors requested exceeds number of variables/columns in data.

fix

Reduce n_factors or increase variables.

AttributeError: 'FactorAnalyzer' object has no attribute 'loadings_'

Called .loadings_ before calling .fit().

fix

Call fa.fit(df) then access fa.loadings_.

RuntimeError: Maximum Likelihood estimation failed to converge. Try increasing max_iter or adjusting tolerance.

ML extraction may not converge with default settings on difficult data.

fix

Set max_iter=1000 or method='principal' (default is 'minres').

Upgrade

Version history

0.5.1latest on PyPI · released Feb 8, 2024

Audit

Dependencies

pandasrequiredDataFrame input

numpyrequirednumerical operations

scikit-learnrequiredPCA/Imputation

scipyrequiredoptimization

Agent activity

28 hits · last 30 days

node

OpenAI (training)

Resources

githubhttps://github.com/EducationalTestingService/factor_analyzer ↗

homepagehttps://github.com/EducationalTestingService/factor_analyzer ↗