Results not invalidated when study data changes

[sgt101]

I removed documents from my canonical directory and then attempted to reanalyse.

Warning:
Sorry, an internal error occurred. Could you please send the authors the text below and a brief note about what you were doing? Thanks!
Traceback (most recent call last):
File "luminoso/window.pyc", line 258, in analyze
File "luminoso/study.pyc", line 367, in analyze
File "luminoso/study.pyc", line 299, in calculate_stats
File "luminoso/study.pyc", line 309, in report_stats
File "luminoso/study.pyc", line 306, in make_info_page
File "luminoso/report.pyc", line 31, in render_info_page
File "jinja2/environment.pyc", line 705, in render
File "<template>", line 17, in top-level template code
File "jinja2/filters.pyc", line 552, in do_round
UndefinedError: 'dict' object has no attribute u'best-solutions-for-me.txt'

(boom head shot!)

best-solutions-for-me.txt was one of the files that I trashed.

[Ken Arnold (kenneth-arnold)]

This behavior is due to cached analysis results data that refers to a canonical document that has been removed. In general, analysis results should be invalidated when any part of the study changes.