The code in print_penn_trees.py shows you how to print parse trees from a Penn treebank file:
import nltk
nltk_reader = nltk.corpus.reader.BracketParseCorpusReader(\
corpus_directory,\
[parse_file])
tree_list = list(nltk_reader.parsed_sents(parse_file))
for ind in sentences:
tree_list[ind].draw()