alright im going about this in the blind so ima throw down my whole task
here and see what ideas you guys throw at me.
I need to take data from a csv file and throw it onto a graph outputting
it in various formats. I need the times in the file to correspond with
the times on the graph, IE 22:30 which would be 10:30pm for those that
dont know military time. I want to pull out and graph each line such as
line1 = blah1 line2 = moreblah line3 = differentblah .... doing this
all with the command line options of things such as start and end dates
to determine what range of the CSV to use. My initial thoughts were to
just program in the blind and see what i came up with, and this is what
i came up with...
- Basically i grep'd out all the data pertaining to Blah1, that is, each
line that had Blah1 in it...and shoved it into a seperate file. Did the
same for blah2 blah3 blah4 ect.. so now i have all these blah files. I
wanted to graph the lines by pulling the data from each of the blah
files, so if line1 was Blah1's data..i could direct it as so. 2 problems
ive run into are as follows
1. this isnt option parsed (optparse)
2. the plots from Blah1's data didnt match up with the times on the
bottom of the graph. per'se in the csv that the time for the selected
dates didnt start at 24:00.
3. trying to perform a string parse to pull the date format out
correctly.
ยทยทยท
###
Im using ruport to do all of this so far but its proving tedious when
ive spent 20 hours a day working on it and getting nowhere fast.
Any ideas to point me in the write direction would be awesome, thanks
fellas!
--
Posted via http://www.ruby-forum.com/.
Michael Linfield wrote:
alright im going about this in the blind so ima throw down my whole task
here and see what ideas you guys throw at me.
I need to take data from a csv file and throw it onto a graph outputting
it in various formats. I need the times in the file to correspond with
the times on the graph, IE 22:30 which would be 10:30pm for those that
dont know military time. I want to pull out and graph each line such as
line1 = blah1 line2 = moreblah line3 = differentblah .... doing this
all with the command line options of things such as start and end dates
to determine what range of the CSV to use. My initial thoughts were to
just program in the blind and see what i came up with, and this is what
i came up with...
- Basically i grep'd out all the data pertaining to Blah1, that is, each
line that had Blah1 in it...and shoved it into a seperate file. Did the
same for blah2 blah3 blah4 ect.. so now i have all these blah files. I
wanted to graph the lines by pulling the data from each of the blah
files, so if line1 was Blah1's data..i could direct it as so. 2 problems
ive run into are as follows
1. this isnt option parsed (optparse)
2. the plots from Blah1's data didnt match up with the times on the
bottom of the graph. per'se in the csv that the time for the selected
dates didnt start at 24:00.
3. trying to perform a string parse to pull the date format out
correctly.
###
Im using ruport to do all of this so far but its proving tedious when
ive spent 20 hours a day working on it and getting nowhere fast.
Any ideas to point me in the write direction would be awesome, thanks
fellas!
I am faced with similar problems. Two words: "relational database". Two
more words: "SQL Cookbook". 
Seriously, though, you probably ought to put your CSV file into MySql or
PostgreSQL first. You *can* do queries/reporting against a raw CSV by
either brute force (read the whole enchilada into RAM if the enchilada
will fit) or making it a database table via ODBC or some other library.
But that's grossly inefficient when modern databases have query planners
designed for just this sort of thing, and modern reporting tools like
ruport interface to them.
By the way, if you have Oracle, DB2 or SQL Server available, they have
some enhanced SQL functions that make this sort of thing easier. If you
don't, bummer.