Test notebook lazy pipeline¶

# Installed packages
import pandas as pd

# Testing
from IPython.utils.capture import capture_output

# Our package
from ydata_profiling import ProfileReport
from ydata_profiling.utils.cache import cache_file

# Read the Titanic Dataset
file_name = cache_file(
    "titanic.csv",
    "https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv",
)
df = pd.read_csv(file_name)

# Generate the Profiling Report (with progress bar)
with capture_output() as out:
    profile = ProfileReport(df, title="Titanic Dataset", progress_bar=True, lazy=False)

assert all(
    any(v in s.data["text/plain"] for v in ["%|", "FloatProgress"]) for s in out.outputs
)
assert len(out.outputs) == 2

# Generate the Profiling Report (without progress bar)
with capture_output() as out:
    profile = df.profile_report(
        title="Titanic Dataset",
        html={"style": {"full_width": True}},
        progress_bar=True,
        lazy=True,
    )

assert len(out.outputs) == 0

with capture_output() as out:
    _ = profile.to_html()


assert all(
    any(v in s.data["text/plain"] for v in ["%|", "FloatProgress"]) for s in out.outputs
)
assert len(out.outputs) == 3

with capture_output() as out:
    _ = profile.to_file("/tmp/tmpfile.html")

assert "Export report to file" in out.outputs[0].data["text/plain"]
assert len(out.outputs) == 1

# Test caching of the iterative building process
with capture_output() as out:
    profile = ProfileReport(df, title="Titanic Dataset", progress_bar=True, lazy=True)
assert len(out.outputs) == 0

with capture_output() as out:
    profile.description_set
assert len(out.outputs) == 1

with capture_output() as out:
    profile.report
assert len(out.outputs) == 1

with capture_output() as out:
    profile.html
assert len(out.outputs) == 1

with capture_output() as out:
    profile.config.html.style.theme = "united"
    profile.invalidate_cache("rendering")
    profile.to_file("/tmp/cache1.html")
assert len(out.outputs) == 2

with capture_output() as out:
    profile.config.pool_size = 1
    profile.html
assert len(out.outputs) == 0

with capture_output() as out:
    profile.config.pool_size = 0
    profile.config.samples.head = 5
    profile.config.samples.tail = 15
    profile.invalidate_cache()
    profile.to_file("/tmp/cache2.html")
assert len(out.outputs) == 4