On 20/05/2007 8:52 PM, Paddy wrote: > On May 20, 2:16 am, John Machin <[EMAIL PROTECTED]> wrote: >> On 19/05/2007 3:14 PM, Paddy wrote: >> >> >> >>> On May 19, 12:07 am, py_genetic <[EMAIL PROTECTED]> wrote: >>>> Hello, >>>> I'm importing large text files of data using csv. I would like to add >>>> some more auto sensing abilities. I'm considing sampling the data >>>> file and doing some fuzzy logic scoring on the attributes (colls in a >>>> data base/ csv file, eg. height weight income etc.) to determine the >>>> most efficient 'type' to convert the attribute coll into for further >>>> processing and efficient storage... >>>> Example row from sampled file data: [ ['8','2.33', 'A', 'BB', 'hello >>>> there' '100,000,000,000'], [next row...] ....] >>>> Aside from a missing attribute designator, we can assume that the same >>>> type of data continues through a coll. For example, a string, int8, >>>> int16, float etc. >>>> 1. What is the most efficient way in python to test weather a string >>>> can be converted into a given numeric type, or left alone if its >>>> really a string like 'A' or 'hello'? Speed is key? Any thoughts? >>>> 2. Is there anything out there already which deals with this issue? >>>> Thanks, >>>> Conor >>> You might try investigating what can generate your data. With luck, >>> it could turn out that the data generator is methodical and column >>> data-types are consistent and easily determined by testing the >>> first or second row. At worst, you will get to know how much you >>> must check for human errors. >> Here you go, Paddy, the following has been generated very methodically; >> what data type is the first column? What is the value in the first >> column of the 6th row likely to be? >> >> "$39,082.00","$123,456.78" >> "$39,113.00","$124,218.10" >> "$39,141.00","$124,973.76" >> "$39,172.00","$125,806.92" >> "$39,202.00","$126,593.21" >> >> N.B. I've kindly given you five lines instead of one or two :-) >> >> Cheers, >> John > > John, > I've had cases where some investigation of the source of the data has > completely removed any ambiguity. I've found that data was generated > from one or two sources and been able to know what every field type is > by just examining a field that I have determined wil tell me the > source program that generated the data.
The source program that produced my sample dataset was Microsoft Excel (or OOo Calc or Gnumeric); it was induced to perform a "save as CSV" operation. Does that help you determine the true nature of the first column? > > I have also found that the flow generating some data is subject to > hand editing so have had to both put in extra checks in my reader, and > on some occasions created specific editors to replace hand edits by > checked assisted hand edits. > I stand by my statement; "Know the source of your data", its less > likely to bite! > My dataset has a known source, and furthermore meets your "lucky" criteria (methodically generated, column type is consistent) -- I'm waiting to hear from you about the "easily determined" part :-) Cheers, John -- http://mail.python.org/mailman/listinfo/python-list