Imprecise Component flagging

[shellover]

When I create a striplog from binary values (i.e. Non-Pay = 0, Pay = 1] and then extract well logs, it always carries over the first value of the next unit.
s = Striplog.from_log(df['Pay'].values, components=comps, basis=df['MD_M'].values)

por = Curve(data=df.por.values, index=df.MD_M.values)

s = s.extract(por.values, basis=df.MD_M.values, name='POR')
so for example if my df looked like this:
MD_M Pay Por
1.0 0 nan
1.5 0 nan
2.0 0 nan
2.5 0 nan
3.0 1 0.3
3.5 1 0.2
4.0 1 0.1
4.5 1 0.25
5.0 0 nan
s[0] would have por values [nan, nan, nan, nan, 0.3]
s[1] would have por values [0.2, 0.1, 0.25, nan]

[rselover]

Updated/tested code to reproduce:

`import matplotlib.pyplot as plt
import pandas as pd
from io import StringIO
from striplog import Striplog, Component
from welly import Curve

csv_string = """MD_M,Pay,Por
1.0, 0, np.nan()
1.5, 0, np.nan()
2.0, 0, np.nan()
2.5, 0, np.nan()
3.0, 1, 0.3
3.5, 1, 0.2
4.0, 1, 0.1
4.5, 1, 0.25
5.0, 0, np.nan()
"""

df=pd.read_csv(StringIO(csv_string))

comps = [
Component({'pay': True}),
Component({'pay': False})
]

s = Striplog.from_log(df['Pay'].values, components=comps, basis=df['MD_M'].values)

por = Curve(data=df.Por.values, index=df.MD_M.values)

s = s.extract(por.values, basis=df.MD_M.values, name='POR')

s[0]

s[1]
`

[rselover]

The problem appears to be with read_at - this produces different results than df.head, offset by approx 1 but actually not systematic

`

[rselover]

continuing pulling the thread, I'm now looking at the behavior of spans

[rselover]

PR 163

#163