Author: Gary Foreman
email: gforema2@illinois.edu
talkbass thread
- matplotlib.pyplot
- numpy
- pyquery
- forecast
- ggplot2
If you wish to run this code make sure monthYearDicts.py, history.py, and
plot.py are in the same directory. history.py imports monthYearDicts.py.
Run by entering ./history.py at the command line, but beware, this code takes
about a half hour to run. ./history.py will store the thread count data as
a numpy array in a file called monthArray.npy, this way, you don't need to
rerun the analysis each time you want to edit the plots. To generate the
plots, run ./plot.py at the command line, and it will create two files
thread_count_[FORUM].png and months_only_[FORUM].png. The thread_count plot
contains the plot of the number of threads posted per month over the history
of the talkbass classifieds. The months_only plot plots the number of basses
sold each month summed over the years 2006-2013.
time_series_bass.R.ipynb is a Jupyter notebook containing R code for a time series analysis of the monthly talkbass counts data. If you wish to run any of the sells in this notebook, you'll need to perform the data scraping first. The data that is presented in the notebook is truncated after January 2016. The plotARIMA.R and plotValidation.R files contain helper function for the Jupyter notebook.