Hello, I have encountered an python interpreter crash when importing pytables (as “import tables”) using newest versions (1.2.2. and 1.2.3).

Older versions up to 1.2 are ok.

 

Platform: winXP Pro service pack 2

Python: 2.4

HDF5: 5-165-win-net

 

I have also built version 1.2.3 from source, but no way.

 

Can you help or fix this?

 

I need some more help: I am using pytables to store timeseries data from wastewater treatment plant monitoring devices and laboratory results for further processing and analysis; the typical table description for lab results is:

 

/Laboratorio/II_0 (Table(8089L,)) 'Ingresso Impianto a valle delle griglie finissime'

  description := {

  "lab": StringCol(length=2, dflt=CharArray(['']), shape=1, pos=0, indexed=False),

  "time": TimeCol(dflt=0.0, shape=1, itemsize=8, pos=1, indexed=False),

  "param": StringCol(length=5, dflt=CharArray(['']), shape=1, pos=2, indexed=False),

  "value": Col(dtype='Float32', shape=1, dflt=0.0, pos=3, indexed=False),

  "text": StringCol(length=50, dflt=CharArray(['']), shape=1, pos=4, indexed=False)}

  byteorder := little

 

I am in need to change, for example, all the records corresponding to a particular ‘time’==t, but only those coming from ‘lab’==’SC’.

I thought that a syntax like:

res = [x for x in lab.II_0 if lab.II_0.cols.time == t and x[‘lab’] == ‘SC’]

would work, but I get a bunch of fake records instead.

 

Thanks a lot,

 

Francesco Tabani

Reply via email to