feat: add the DraftValidation class

[rich-iannone]

This adds DraftValidation, a class that can generate a validation plan based on any table that Pointblank supports. Example usage:

import pointblank as pb

nycflights = pb.load_dataset("nycflights", tbl_type="duckdb")

pb.DraftValidation(data=nycflights, model="anthropic:claude-3-5-sonnet-latest")

This generates:

import pointblank as pb

# Define schema based on column names and dtypes
schema = pb.Schema(columns=[
    ("year", "int64"), ("month", "int64"), ("day", "int64"),
    ("dep_time", "int64"), ("sched_dep_time", "int64"), ("dep_delay", "int64"),
    ("arr_time", "int64"), ("sched_arr_time", "int64"), ("arr_delay", "int64"),
    ("carrier", "string"), ("flight", "int64"), ("tailnum", "string"),
    ("origin", "string"), ("dest", "string"), ("air_time", "int64"),
    ("distance", "int64"), ("hour", "int64"), ("minute", "int64")
])

# The validation plan
validation = (
    pb.Validate(
        data=your_data,
        label="Draft Validation",
        thresholds=pb.Thresholds(warn_at=0.10, stop_at=0.25, notify_at=0.35)
    )
    .col_schema_match(schema=schema)
    .row_count_match(count=336776)
    .col_count_match(count=18)
    .rows_distinct()
    .col_vals_not_null(columns=[
        "year", "month", "day", "sched_dep_time", "sched_arr_time", 
        "carrier", "flight", "origin", "dest", "distance", "hour", "minute"
    ])
    .col_vals_eq(columns="year", value=2013)
    .col_vals_between(columns="month", left=1, right=12)
    .col_vals_between(columns="day", left=1, right=31) 
    .col_vals_between(columns="dep_time", left=1, right=2400, na_pass=True)
    .col_vals_between(columns="dep_delay", left=-43, right=1301, na_pass=True)
    .col_vals_between(columns="arr_time", left=1, right=2400, na_pass=True)
    .col_vals_between(columns="air_time", left=20, right=695, na_pass=True)
    .col_vals_between(columns="hour", left=1, right=23)
    .col_vals_between(columns="minute", left=0, right=59)
    .col_vals_gt(columns="distance", value=16)
    .col_vals_in_set(columns="origin", set=["EWR", "LGA", "JFK"])
    .interrogate()
)

validation

which, when run, is fully working and produces no failing test units:

Currently, this is lean on support for model providers (only Anthropic and OpenAI for now).

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 93.93939% with 2 lines in your changes missing coverage. Please review.

Project coverage is 99.16%. Comparing base (39cc8b2) to head (377390b).

Files with missing lines	Patch %	Lines
pointblank/draft.py	92.59%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #60      +/-   ##
==========================================
- Coverage   99.22%   99.16%   -0.06%     
==========================================
  Files          13       14       +1     
  Lines        2974     3003      +29     
==========================================
+ Hits         2951     2978      +27     
- Misses         23       25       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.