I am using an API that returns what appears to be a CSV string that i need to parse for two decimal numbers and then need to append those numbers to separate lists as decimal numbers (also while ignoring the timestamp at the end):
returned_string_from_API = '0,F,F,1.139520,1.139720,0,0,20160608163132000'
decimal_lowest_in_string = []
decimal_highest_in_string = []
Processing time is a factor in this situation so, what is the fastest way to accomplish this?
Split the string by comma:
>>> string_values = returned_string_from_API.split(',')
>>> string_values
['0', 'F', 'F', '1.139520', '1.139720', '0', '0', '20160608163132000']
Get the values from string:
>>> string_values[3:5]
['1.139520', '1.139720']
Convert to float:
>>> decimal_values = [float(val) for val in string_values[3:5]]
>>> decimal_values
[1.13952, 1.13972]
Get min and max in the appropriate list:
>>> decimal_lowest_in_string = []
>>> decimal_highest_in_string = []
>>> decimal_lowest_in_string.append(min(decimal_values))
>>> decimal_lowest_in_string
[1.13952]
>>> decimal_highest_in_string.append(max(decimal_values))
>>> decimal_highest_in_string
[1.13972]
1.139520,1.139720 part and split it. Thus not having to process the whole input.1) The version which does not rely on cvs
returned_string_from_API = '0,F,F,1.139520,1.139720,0,0,20160608163132000'
def isfloat(value):
try:
float(value)
return True
except ValueError:
return False
float_numbers = filter(isfloat, returned_string_from_API.split(','))
2) try pandas package
Fastest way is to use regular expression. Readability is another issue..
import re
returned_string_from_API = '0,F,F,1.139520,1.139720,0,0,20160608163132000'
decimal_lowest_in_string = []
decimal_highest_in_string = []
re_check = re.compile(r"[0-9]+\.\d*")
m = re_check.findall(returned_string_from_API)
decimal_lowest_in_string.append(min(m))
decimal_highest_in_string.append(max(m))
str.split(',')