Yesterday, I worked for about a half hour on generating the historical report.
On Monday, I tried to index the central dataframe using a MultiIndex (timestamp, room #) so that I could select data by time, but encountered some KeyErrors when I tried to actually get a specific row. In the debugging process yesterday, I took a step back, created a test DataFrame, and tried to index it. This helped me realize my very simple mistake -- that I should have been using .loc instead of just appending brackets to the name of the DataFrame!
Additionally, I started planning out exactly how I was going to select a week of data. The most likely option seems to be taking the desired start time, adding one week to it, and getting all of the data in between. (I had considered using some kind of for loop to separate out each interval, but some weeks may be missing some intervals due to days off from school or other issues. Therefore, I decided to separate out each interval after selecting the initial week of data.)
Yesterday, I worked for around an hour on the new version of the Weekly Report.
I mainly focused on trying to generate a report from historical data (as compared to taking a week to generate each new report). This would be helpful because it would further the application of separation of concerns; the part of the program that analyzes the data would be able to run immediately on any week of data. This would allow thresholds to be changed easily, to create a different report from the same data.
As I began this new version, I wrote psuedocode, and documented issues I encountered along the way, so I could check back on them later.
Right now, my document looks like this:
I just realized that I never updated this blog with news of my 2 presentations!
First, I presented my work to the Andover Green Advisory Board. This was an exciting experience for me; talking about my work to an audience for the first time was amazing, especially when they were so engaged and helpful!
A week later, I presented to the Facility Department. For this presentation, I went more in-depth into what the report can offer as a workable product. I was thrilled with the response, as the members began discussing different use cases and options to add to the report's functionality.
In between all this, I have had a few issues with the report itself. (For the last couple weeks, it seems to have been running on the same week of data.) Now that we are on February Break at school, I can get to work on debugging this issue and any others that arise. I also want to create a version that operates entirely on historical data.
I am so grateful to AGAB and to the Facility for giving me these amazing opportunities to present!