2008-03-21

PR: 21 March 2008

First slippage, which seems odd considering that I haven't built up a long term plan for the project. Hmm. Something to add to the To-Do list.


Work In Progress: Project Relstring


Last time looked at: 20 Mar 08

Last actual wordage added: --

Words added 21 Mar 08: 523

Current word count: 523

Estimated final word count: Need to target journal/conference

Deadline: Depends on target. Earliest is 9 May 08

Comments: Ignored RedHat, except to make a decision on which upgrade approach to take. I did organize my conception of upcoming tasks in my mindmap. This will help with time allocation for short-term planning. The 523 words come from the design "document" that I put together for my Top-K Difference Analyzer. This was a greater level of detail than I've been putting into my designs for grad school coding projects, and it forced me to think through errors in my conception of the project. On the down-side, I didn't get the design done on the randomizer.

Tomorrow: I've added the long-range plan to the mindmap, it is in the general mix of things to get done. I need to read this article by Abraham Bookstein that was recommended to me by Paul Thompson. Off project, I'll be trying to get the randomization test design done.

20 March 2008

Not such a bad day. Didn't get going as early as I'd hoped. So, working late again.


Work In Progress: Project Relstring


Last time looked at: 20 Mar 08

Last actual wordage added: --

Words added 20 Mar 08: --

Current word count: --

Estimated final word count: Need to target journal/conference

Deadline: Depends on target. Earliest is 9 May 08

Comments: Started research on RedHat upgrades, and e-mailed my advisor with two different options. I've got two (three?) boxes that I need to take from Fedora Core 6 up to Fedora 8. I got the coding written to implement the features. Each script consists of 100 updates. The last time I ran one of these scripts, it only seemed to take a few hours. The current one looks like it will take a bit longer than that.

Tomorrow: I need to put together a to-do list. I've used FreeMind to put together broad task-type categories, and then put a branch in each of those categories for each project. While I'm at it, I should have the design and framework for generating distance features for my project. I'll be looking at the measures described in Fagin, et al. Off project, I'll be looking to put together the code framework for implementing a randomization test as an alternative to the Wilcoxon signed rank tests that I've been using. The idea is that I can reuse that code for implementing other tests.

2008-03-20

19 March 2008

I've been meaning to start this up. The fantasy writer Lawrence Watt-Evans posts regular (daily) progress reports for works in progress. It is a way of tracking goal progress, which motivates some. Others, experts in dissertation writing, have noted that making public ones progress goals can also help motivate forward progress (and if I was a good Academic, I'd have a cite for that). So, while I modify LWE's format


Work In Progress: Project Relstring


Progress report:

Last time looked at: 19 Mar 08

Last actual wordage added: --

Words added 19 Mar 08: --

Current word count: --

Estimated final word count: Need to target journal/conference

Deadline: Depends on target. Earliest is 9 May 08

Comments: I've finished creating non-weighted features for my dataset. I needed to use a batch process to update the database, something I should probably do automatically. Also learned about enabling user web pages on a Red Hat box.

Tomorrow: Minimum, I need to add the weighted features and weighted average-features that include all runs. Research RedHat upgrade procedures.