Add quad gating to rushd

[meson800]

A common flow task is quad gating. I think this example image is probably tool-generated and not done in code, but it is a good example:

Unlike that image, we can also make it more useful if we can show gating per colored condition. However, making a plot with annotations like this is typically at least slightly annoying since it involves a subsequent pandas groupby. We can make a helper function that does a couple things:

1. Makes a scatter or kdeplot using the seaborn function.
2. Does a groupby (by whatever the hue column is, if it exists).
3. Calculates the gated percentage by the hue column.
4. Draws axvline / axhline at certain gate values and adds annotations in the corners. If there are multiple hues, add multicolored text so you can see e.g. the black population has 25% in this quadrant and the purple has 45%.

Implementation tips

The easiest way to do this is probably as a decorator that captures arguments. I think a clean way to implement this would be something that looks like this type of call:

rd.flow.quad_gated(sns.kdeplot)(data=df, x='mRuby2-A', y='mGL-A', x_gate=500, y_gate=2500, hue='condition', log_scale=True)

How do we implement such a thing? A decorator is a function that takes a function and returns a function that does what you want. It probably looks like:

# In rd.flow
def quad_gated(sns_f):
  def decorated(*args, **kwargs):
    # Args is the list of positional arguments, kwargs is the keyword arguments.

    # Start by pulling out the "x_gate" and "y_gate" kwargs
    x_gate = kwargs['x_gate']
    y_gate = kwargs['y_gate']
    # strip these keywords out so the seaborn function doesn't get confused
    del kwargs['x_gate']
    del kwargs['y_gate']

    # Extract the dataframe / x col / y col / hue col variables
    df = args[0] if len(args) > 0 else kwargs['data']
    x_col = kwargs['x']
    y_col = kwargs['y']
    hue_col = kwargs['hue'] if 'hue' in kwargs else None

    # call the seaborn function
    g = sns_f(*args, **kwargs)
    # g is now the matplotlib Axes object
    
    # TODO: do the groupby gating on df to get percentages
    # TODO: pull out the hue mapping from the g object
    # TODO: add axvline, axhline, and annotations in the corners
return decorated

I assigned Emma since it looks like you use this type of gate (and can contribute to rushd!), Kasey because you've added stuff before.