ENH: add from_edges to IntervalIndex

[Xiaoven]

Feature Type

• Adding new functionality to pandas

• Changing existing functionality in pandas

• Removing existing functionality in pandas

Problem Description

I wish I could use pandas to create an IntervalIndex like [0.0, 0.1), [0.1, 0.2), ..., [0.9, 1.0].

Currently, I successfully cut my data frame using the following code:

bin_edges = [i / 10.0 for i in range(0, 11)]  # Create bins from 0.0 to 1.0 with a size of 0.1
df['bin'] = pd.cut(df['ratio'], bins=bin_edges, include_lowest=True)

print(df['bin'].unique())
# Output:
# [(-0.001, 0.1], (0.9, 1.0], (0.4, 0.5], (0.8, 0.9], (0.3, 0.4], ..., (0.5, 0.6], (0.2, 0.3], (0.1, 0.2], (0.6, 0.7], NaN]
# Length: 11
# Categories (10, interval[float64, right]): [(-0.001, 0.1] < (0.1, 0.2] < (0.2, 0.3] < (0.3, 0.4] ... (0.6, 0.7] < (0.7, 0.8] < (0.8, 0.9] < (0.9, 1.0]]

Then, I want to reindexing to add back the missing missing bins and fills them with zero counts using fillna(0) with the following code

bin_counts = df['bin'].value_counts().reindex(pd.IntervalIndex.from_breaks(bin_edges)).fillna(0)

However, I found the (-0.001, 0.1] with a non-zero count is replaced by (0, 0.1] with a zero count.

I ask ChatGPT, and it recommends to use bin_counts = df['bin'].value_counts().reindex(pd.IntervalIndex.from_edges(bin_edges)).fillna(0) which lead to AttributeError: type object 'IntervalIndex' has no attribute 'from_edges'

Feature Description

Add a new function to IntervalIndex, from_edges, to return IntervalIndex based on bin edges.

 def from_breaks(cls, bins) -> IntervalIndex

Alternative Solutions

I haven't found one yet

Additional Context

No response

[mroeschke]

Thanks for the suggestion but I believe you want from_breaks https://pandas.pydata.org/docs/reference/api/pandas.IntervalIndex.from_breaks.html#pandas.IntervalIndex.from_breaks so closing

[Xiaoven]

Thanks for the suggestion but I believe you want from_breaks https://pandas.pydata.org/docs/reference/api/pandas.IntervalIndex.from_breaks.html#pandas.IntervalIndex.from_breaks so closing

from_breaks does not fully meet my requirements for creating intervals. My expected intervals are [0, 0.1), [0.1, 0.2), ..., [0.8, 0.9), [0.9, 1.0], with both 0 and 1 included in the range. To achieve this with pd.cut, I use pd.cut(df['ratio'], bins=bin_edges, include_lowest=True) to treat the first interval specially as (-0.001, 0.1], ensuring that 0 is included.

However, when I attempt to use pd.IntervalIndex.from_breaks(bin_edges, closed='left'), the last interval becomes [0.9, 1.0), which excludes 1.0.

When I try pd.IntervalIndex.from_breaks(bin_edges, closed='right'), the first interval is (0.0, 0.1], which excludes 0.0.

Finally, using pd.IntervalIndex.from_breaks(bin_edges, closed='both') includes both 0 and 1 in the intervals, but this approach leads to overlapping intervals, which is not what I want.

Given these issues, I'm looking for a solution that allows me to include both 0 and 1 in the intervals without overlap. Is there an alternative method or adjustment that I can apply to from_breaks to achieve this? What about adding something like include_lowest to from_breaks?