'numexpr' - how to find a pandas supported version?

[SophMC]

@aaren
I believe the pandas function query() could be very useful for me, but when I try to use it, I get this error
wind.query('year' > 1986 and 'year' < 1999)

ImportError: 'numexpr' is not installed or an unsupported version. Cannot use engine='numexpr' for query/eval if 'numexpr' is not installed

The documentation here says:
"numexpr: for accelerating certain numerical operations. numexpr uses multiple cores as well as smart chunking and caching to achieve large speedups. If installed, must be Version 2.1 or higher (excluding a buggy 2.4.4). Version 2.4.6 or higher is highly recommended."

It IS installed - version 2.5.

So, is pandas not trained to use 2.5? I believe there are some clues in this discussion, but I don't follow what they are saying.

Should I install a different version of numexpr? or accept that pd.query() is beyond my grasp :(

p.s. I updated my whole anaconda distribution, but I still only have 0.18.0. I think 0.18.1 was only released on 3rd May 2016.

[SophMC]

I've also tried adding parser = 'python', engine = 'python, engine = 'numexpr' but to no avail:

wind.query('year' > 1986 and 'year' < 1999, parser='python')

[aaren]

If you do wind.query('year' > 1986, engine='python') what error do you get?

[SophMC]

Very strange. I'm not getting the numexpr ImportError at all now.
To answer your question:
wind.query('year' > 1986, engine='python') #KeyError: True

Some other variations:
wind.query('year' > 1986 or 'year' < 1999)# KeyError: True

`wind.query('year' > 1986 and 'year' < 1999)# KeyError: False``

wind.query('year' > 1986 or 'year' < 1999, engine='python')# KeyError: True

The documentation here suggests that the following two should give the same result. But they don't.
wind.query('year' > 1986) #KeyError: True
wind[wind.year > 1986]
date_time year month day hour ws
2593 1987-01-01 06:00:00 1987 1 1 6 5.82639
2594 1987-01-01 12:00:00 1987 1 1 12 7.30782
2595 1987-01-01 18:00:00 1987 1 1 18 5.05220
2596 1987-01-02 06:00:00 1987 1 2 6 5.23692
2597 1987-01-02 12:00:00 1987 1 2 12 6.36581
2598 1987-01-02 18:00:00 1987 1 2 18 4.72313
2599 1987-01-03 06:00:00 1987 1 3 6 3.67479
2600 1987-01-03 18:00:00 1987 1 3 18 3.77146
2601 1987-01-04 06:00:00 1987 1 4 6 4.40437
2602 1987-01-04 18:00:00 1987 1 4 18 4.06409

[aaren]

Your expression needs to be a string.

wind.query("year > 1986 or year < 1999")

[SophMC]

Thank you. I still don't understand what happened with ImportError at the beginning, but it doesn't matter now!