Task: Build data set for a logistic regression model that predicts the state of a protein

Data location:
super_secret_md.tar.gz

Requirements:
- Data set must be in a CSV file format.
- The CSV must not have any column headers and the last column(farthest right) must contain the label.
- Within the parent data directory 'super_secret_md' are two directories named 'closers' and 'openers'. These will be the
  class labels you will be predicting.
- Any data within the closers directory will have a label of 1 and anything within the openers directory a label of 0.
- Within each of these 2 directories, there will be additional directories that that contain several .dat data files.
  These data files will have our feature data in them.
- One .dat file == one sample
- Each file will be formatted as 2 values per row. The first value can be ignored but the second
  will be the feature used in the data set.
- If the contains NaN values, handle it as you see fit.
- The max length of the feature vectors is 20. Fill with zeros to make shapes consistent.
- Feel free to install Python modules.
- To run the model script, python analysis.py path/to/data.csv