CODEWITHSUNDEEP
pandaspython-3.6
steff
steff

Reputation: 906

pandas df lookup value based on multiple conditions

my df called symbols looks like this:

                                   local_symbol            code globex expiry_date type  strike
symbol                                                                                         
OZN20121221P00014900_FOP  P OZN DEC 12    14900  CBT_21_F2013_S    OZN  2012-12-21    P   149.0
OZN20121221C00012500_FOP  C OZN DEC 12    12500  CBT_21_F2013_S    OZN  2012-12-21    C   125.0
OZN20121221P00012450_FOP  P OZN DEC 12    12450  CBT_21_F2013_S    OZN  2012-12-21    P   124.5
OZN20121221C00013900_FOP  C OZN DEC 12    13900  CBT_21_F2013_S    OZN  2012-12-21    C   139.0
OZN20121221C00010700_FOP  C OZN DEC 12    10700  CBT_21_F2013_S    OZN  2012-12-21    C   107.0

using pandas 0.22.0 the following worked:

exp_date = dt.date(2012, 12, 21)
code = 'CBT_21_F2013_S'
type = 'P'
strike = 124.5
symbols.loc[(symbols.expiry_date == exp_date)
            & (symbols.code == code)
            & (symbols.type == type)
            & (symbols.strike == strike)]

and returned OZN20121221P00012450_FOP as the expected value. In Pandas 0.23 i get an empty dataframe. Your help would be greatly appreciated.

Upvotes: 1

Views: 1905

Answers (1)

jezrael
jezrael

Reputation: 863256

I believe need convert column expiry_date to dates and for compare floats use numpy.isclose:

symbols.loc[(symbols.expiry_date.dt.date == exp_date)
            & (symbols.code == code)
            & (symbols.type == type)
            & np.isclose((symbols.strike, strike)]

Upvotes: 1

Related Questions