Julia新手咨询：如何用searchsortedfirst查找指定y字段的Record元素

Hey there! Since you're new to Julia, let's walk through how to solve this problem clearly and simply.

First, let's recap your goal: you have a sorted array of Record objects, and you want to use searchsortedfirst to find the index of the first Record where the y field equals a given x.

Do you need by, lt, or both? #

You only need the by transformation function here—no need to define a custom lt comparator. Here's why:

• The by parameter tells searchsortedfirst which part of your Record to use for comparison. In your case, we care about the y field, so we'll pass a function that extracts y from each Record.
• The default lt (less-than) comparator (isless) works perfectly for integer values like your y::Int64, since it follows the standard numeric ordering that searchsortedfirst expects.

Example Code #

Let's put this into practice. Note that in Julia 1.0+, we use struct instead of the old type keyword (I'll adjust that for you to follow modern syntax):

# Define your Record type (modern Julia syntax)
struct Record
    y::Int64
    value::Float64
end

# Create a sorted array of Records (make sure it's sorted by y!)
A = [Record(1, 0.5), Record(3, 1.2), Record(5, 2.0), Record(7, 3.1)]

# The x value we want to match against the y field
target_y = 3

# Call searchsortedfirst with the by parameter
idx = searchsortedfirst(A, target_y, by=r -> r.y)

# Verify we found a matching element (since searchsortedfirst returns the first >= target)
if idx <= length(A) && A[idx].y == target_y
    println("Found matching Record at index $idx: $(A[idx])")
else
    println("No Record with y = $target_y exists in the array")
end

What's happening here? #

1. The by=r -> r.y argument tells searchsortedfirst to ignore the value field and only compare the y values of your Records.
2. searchsortedfirst returns the first index where the y value is greater than or equal to target_y. Since your array is sorted, if a Record with y == target_y exists, this will be its index (or the first occurrence if there are duplicates).
3. We add the check A[idx].y == target_y to handle edge cases: when target_y is larger than all y values in the array (which would make idx = length(A)+1, out of bounds) or when the first ">= element" is actually larger than target_y.

If your array was sorted in descending order of y, you'd just add rev=true to the call:

idx = searchsortedfirst(A, target_y, by=r -> r.y, rev=true)

That's all you need! No custom lt function required here—keep it simple with the by parameter.

内容的提问来源于stack exchange，提问作者 J.Kennsy