The tool is available here.
The format for target files is:
line: targetID(integer),priorStartWeek(count from 1),releaseWeek(start of posterior),posteriorEndWeek,[comma separated list of floats for the vector]
Each target goes on a new line.
Download the sample target file here: targets.csv
The format for control files is:
line: [comma separated list of floats - week 1 for each control]
line: [comma separated list of floats - week 2 for each control]
|
V
Each control spans every line; each week goes on a new line.
Download the sample controls file here: controls.csv
Both samples are available in this zip: ciraSamples.zip
The following questionnaire was sent to developers who expressed an interest in receiving a report on their significant release.
Answering these questions will take only 12 keystrokes, including 6 return key presses.
However, if you want to elaborate on your answer, please do feel free to do so in free text.
We are very interested to hear your views on whether this research could be useful to you.
1. Did you find the time series report useful? 0 - no, 1 - yes
2. Would you be interested in further time series reports on your app(s)? 0 - no, 1 - yes
3. Do you agree that the release impacted your app's performance? 0 - no, 1 - yes
4. Are you aware of any external influences, such as advertising campaigns, that could have led to the observed effect? 0 - no, 1 - yes
5. Would you be interested in learning about potential contributing factors to significant releases? 0 - no, 1 - yes
6. Would you make any changes to your app releasing strategy based on these findings? 0 - no, 1 - yes
Read the results from our additional runs using 10, 20, 50, 200, 500 and 1000 topics in this report: FSE16_Topics.pdf
w.martin (at) ucl.ac.uk