Monday:
I spent most of my morning working on researching how to make an installer file. This was inspired by a meeting I had about integrating my projects into the workflow at CMI. If we could set the file up in the same to place on every computer we could devise an R script that would make a command-line call to my program. This would make it easy and efficient to automate and integrate my work. I was trying to use the default app provided for making the files by Microsoft. They did not work well so I downloaded a third-party app that would allow me to make the app install correctly. I did a number of trial tests and liked the results. I planned on making final products tomorrow.
Tuesday:
I created a new installer file that I thought would be the final version. I shared this with my fellow employees and had them run some tests. It turns out there was an error with one of the higher-end versions where the arguments were not reading incorrectly. I had not saved many of my previous versions and had no point of reference to fix other errors I found when debugging. This took me the rest of the day to debug.
Wednesday:
I finished debugging and had everyone run tests again. We found no more errors and I spent the rest of the day making training data for one of our AI programs
Thursday:
We found another error. The multicore version was reading out the wrong number of files. I figured out each core skipped the first file. I then did the same thing as the day before.
There are no comments published yet.