High false positive rate

Python static analysis with pylint has a fundamental problem of a high rate of false positive findings. It does not plague all checkers, but especially important ones like the type checker that emits e.g. the no-member error (E1101). Given the dynamic nature of python and the meta-programming capabilities used by many python libraries, this is not surprising.

Missing dependencies lead to false negatives

Analyzing python code in isolation from its dependencies only makes sense for some checks. Especially the type-based checks require knowledge of called interfaces. Unfortunately, most other hosted analysis solutions do not install dependencies. PyCodeQual on the other hand, detects and installs dependencies. Read more about this here.