https://github.com/kalpanabaheti/gamesdau
https://github.com/kalpanabaheti/gamesdau
Last synced: 8 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/kalpanabaheti/gamesdau
- Owner: kalpanabaheti
- Created: 2023-10-06T04:50:04.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-10-23T12:21:36.000Z (about 2 years ago)
- Last Synced: 2025-01-07T14:12:11.476Z (10 months ago)
- Language: Python
- Size: 135 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# GamesDAU
## Summary of .py files -
**1. *dataloading*:** Loading CSV, viewing it via DataFrames, and extracting iOS and Android sub-datasets.
**2. *initialplotting*:** Methods for visualising the trends of multiple cohorts on a graph, and consequently visualising a full year of cohorts.
**3. *timeseriesanalysis*:** Methods developed for accurately identifying all significant junctures of change in DAU across cohorts
**4. *exceptionregionanalysis*:** Methods for understanding regions of ambiguity better using summary statistics.
**5. *exceptionregioninference*:** Assuming a linear curve fit (which can be switched to other approaches such as case-wise Bayesian perturbation, or other forms of regression, etc.), inferring a model to explain the exception region.
**6. *generalinference*:** Time-series based curve-fitting (exponential smoothing, to start with) on general trend of major alert regions per 60-day-runs across cohorts. This may also be used to extrapolate preliminary baseline for future data (predictive mode).
**7. *correlationDAU*:** Methods for assessing correlation between truncated previous summative DAU population and current cohort's Day 1 DAU.
## Game Intelligence Official Report
**Link:** https://docs.google.com/presentation/d/10Ujh1dya4eUI_rUDGwhAjOecFQU2nhaHd1oHLCULJXI/edit?usp=sharing