As usual, haven't posted in a long time. Never found much to talk about, but nowadays I find myself looking at problems and inconveniences in my life, and others', and thinking of ways to solve them.
Example. With the abolishing of rickshaws from the main road near leading up to New Market and Nilkhet, the road in front of New Market has become more jammed than ever with parked cars and stationary rickshaws. Right now it is a two-way street, with two lanes on each side and a lane for parking cars on. A simple way to solve the jam would be to allow only cars on the side further away from NM, and only rickshaws on the side closer to it. Sure, cars would have to exit through the other, further side on their way out, but then, that's what they're there for.
Work
Started at One Bank in the beginning of December. Worked, or observed, my way through a lot of stuff but I've finally seen what to me is the most interesting part of it all: the raw data generated by the computer system of the bank's daily activities. These data are in the form of plain text files arranged into folders, essentially by date. They are just crying to be pulled in and processed programmatically by Excel or some such program. For example, there are daily data files about fixed deposits which mature on the day; and new accounts (including loans) which were opened on the previous day.
In the new accounts example, the information in each file (each day's report) includes a grouping of accounts by branch, count of new accounts in each branch, and detailed information on each account (one account per line). The way it is arranged makes it possible to parse it and pull out the most useful data -- for example, the count of new accounts opened in each branch. If one does this every day to keep current, one can graph the daily account opening activity for each branch, and what's more, put these graphs together into a combined `3-D' graph for ease of comparison. This gives, over time, a nice high-level view of account activity throughout the bank.
This is exactly what am now trying to do with Excel and a well-crafted macro at the bank. Have made some progress, and think the parsing bit is taken care of thanks to Excel's, well, excellent plain text file importing/parsing capabilities. But a lot of it is still left, including programmatically generating pivot tables and charts for new months. Should be quite a challenge. If they let me do this, even intermittently, it should make it very interesting at work. Don't know who it will really help, though, to be realistic. At this point it's just a shiny toy, a very high-level view which branch employees may not find ultimately useful and thus may lose interest in rapidly.
But still look forward to exploring more of the daily reports and perhaps even getting something useful out of them.
No comments:
Post a Comment