pca.py: Add ability to drop columns (questions) from input data

diff --git a/posts/Factor_analysis/pca.py b/posts/Factor_analysis/pca.py
index 03d1639690e0a6e69acea6e808c65fdeabd797ba..0e8aaaf2795a36ddddb425bae55718caeea11501 100755 (executable)
--- a/posts/Factor_analysis/pca.py
+++ b/posts/Factor_analysis/pca.py
@@ -597,6 +597,9 @@ if __name__ == '__main__':
     parser.add_argument(
         '-u', dest='unitary', default=False, action='store_const', const=True,
         help='use unitary eigenvalues for factor loading')
     parser.add_argument(
         '-u', dest='unitary', default=False, action='store_const', const=True,
         help='use unitary eigenvalues for factor loading')

+    parser.add_argument(
+        '-d', dest='drop', type=int, action='append',
+        help='drop a column from the analysis')

     parser.add_argument(
         'path',
         help='path to the tab-delimited data file to analyze')
     parser.add_argument(
         'path',
         help='path to the tab-delimited data file to analyze')


@@ -604,4 +607,6 @@ if __name__ == '__main__':
     args = parser.parse_args()
 
     scores = _numpy.genfromtxt(fname=args.path, delimiter='\t')
     args = parser.parse_args()
 
     scores = _numpy.genfromtxt(fname=args.path, delimiter='\t')

+    if args.drop:
+        scores = _numpy.delete(scores, args.drop, axis=1)

     analyze(scores=scores, output_dim=args.output_dim, unitary=args.unitary)
     analyze(scores=scores, output_dim=args.output_dim, unitary=args.unitary)