Proposed ideas for updating the Tea API for V1

tea_oop_api_proposals.py

# Goal: Provide a more OOP API that enables end-users to have multiple analyses simultaneously

# Current V0 API
    variables = [
        {"name": "Var0", "data type": "interval"}
        {"name": "Var1", "data type": "ratio",},
    ]
    experimental_design = {
        "study type": "observational study",
        "contributor variables": "Var0",
        "outcome variables": "Var1",
    }
    assumptions = {"Type I (False Positive) Error Rate": 0.05, 
                   "normal distribution": ["Var1"]}
    # Provide data
    tea.data("data.csv")
    # Defie variables
    tea.define_variables([
        {"name": "Var0", "data type": "interval"}
        {"name": "Var1", "data type": "ratio",},
    ]) # equivalent to tea.define_variables(variables)
    # Define study design 
    tea.define_study_design({
        "study type": "observational study",
        "contributor variables": "Var0",
        "outcome variables": "Var1",
    }) # equivalent to tea.define_study_design(experimental_design)
    # Assume
    tea.assume({"Type I (False Positive) Error Rate": 0.05, 
                   "normal distribution": ["Var1"]}) # equivalent to tea.assume(assumptions)
    # Hypothesize, Execute, & obtain results
    results = tea.hypothesize(["Var0", "Var1"], ["Var0~V1"]) 
    
# V1 API
# Provide data at construction
tea_obj = tea.Tea("data.csv") 
# OR provide data later
tea_obj = tea.Tea()
tea_obj.add_data("data.csv")
# Declare var one at a time
tea_obj.declare_vars(name="Var0", type="interval")
tea_obj.declare_vars(name="Var1", type="ratio")
# OR all at once
tea_obj.declare_vars(name=["Var0", "Var1"], type=["interval", "ratio"])
# Specify study design
tea_obj.specify_design(type="observational study", xs=["Var0"], ys=["Var1"])
# Assume
tea_obj.assume(property="Type I (False Positive) Error Rate", value=0.05)
tea_obj.assume(property="normal distribution", var="Var1")
# Hypothesize
tea_obj.hypothesize(hypothesis="A>B",...) # String hypothesis could be syntactic sugar or kwarg 
# Execute
tea_obj.brew() # provides output

"""
Notes
- declare_vars = would ideally infer types for dataset and allow end-user to override defaults/inferred types
- hypothesize = adds/keep track of hypotheses
- brew = executes valid statistical tests for all hypotheses, applying multiple comparisons when there is more than one hypothesis
"""