commit 6c09da6ddd72d3d29f7045d040dfb8ce15b6a075
parent 347be5310855537b93d034bbd5338ce6b5afeca9
Author: Jared Tobin <jared@jtobin.io>
Date: Wed, 3 Jun 2026 15:45:13 -0230
meta: refine text
Diffstat:
3 files changed, 13 insertions(+), 27 deletions(-)
diff --git a/CHANGELOG b/CHANGELOG
@@ -1,6 +1,6 @@
# Changelog
-- 0.1.0 (2026-05-31)
- * Initial release. Anytime-valid sequential testing via
+- 0.1.0 (2026-06-03)
+ * Initial release, supporting anytime-valid sequential testing via
e-processes: bounded-mean and paired two-sample tests, with
fixed-lambda, aGRAPA, and ONS bettors.
diff --git a/README.md b/README.md
@@ -1,4 +1,4 @@
-# eproc
+# ppad-eproc
[](https://hackage.haskell.org/package/ppad-eproc)
@@ -6,47 +6,38 @@
Anytime-valid sequential hypothesis testing for bounded random
variables, via the e-process / betting framework of
-[Waudby-Smith & Ramdas (2024)][wsr24]. Bounded-mean and paired
-two-sample tests are valid under optional stopping: reject as soon as
-the wealth process exceeds `1/alpha`, with type-I error controlled at
-`alpha` regardless of when the user stops streaming samples.
+[Waudby-Smith & Ramdas (2024)][wsr24].
## Usage
A sample GHCi session:
```
- > -- import qualified
> import qualified Numeric.Eproc.Bounded as Bounded
>
- > -- test H_0: E[X] = 0.5 for samples in [0, 1] at alpha = 1e-3,
+ > -- hypothesis: E[X] = 0.5 for samples in [0, 1] at alpha = 1e-3, tested
> -- with the ONS bettor
> let cfg = Bounded.config 0.5 0.0 1.0 1.0e-3 Bounded.Ons
+ > let s0 = Bounded.initial cfg
>
- > -- streaming interface: 'initial' then fold observations through 'update'
- > let s0 = Bounded.initial cfg
- > let xs = [1, 1, 0, 1, 1, 0, 1, 1, 1, 1] -- mean 0.8, drifts from H_0
- > let s10 = foldl (Bounded.update cfg) s0 xs
+ > -- ten observations (drifting from hypothesis), and state afterwards
+ > let xs = [1, 1, 0, 1, 1, 0, 1, 1, 1, 1]
+ > let s10 = foldl' (Bounded.update cfg) s0 xs
>
- > -- inspect wealth and verdict at any point
- > Bounded.samples s10
- 10
+ > -- inspect wealth and stopping decision at any point
> Bounded.log_wealth s10
0.7182493502552663
> Bounded.decide cfg s10
Continue
>
- > -- with enough evidence the test rejects
- > let s300 = foldl (Bounded.update cfg) s0 (concat (replicate 30 xs))
+ > -- with enough evidence, the hypothesis is rejected
+ > let s300 = foldl' (Bounded.update cfg) s0 (concat (replicate 30 xs))
> Bounded.log_wealth s300
53.092214534054165
> Bounded.decide cfg s300
Reject
```
-For the paired two-sample mean-equality test, see
-`Numeric.Eproc.Paired`.
-
## Documentation
Haddocks (API documentation, etc.) are hosted at
@@ -86,10 +77,6 @@ Current benchmark figures on an M4 Silicon MacBook Air look like (use
std dev 46.74 ns (34.00 ns .. 64.63 ns)
```
-The inner update loop is fully fused: the `fixed` bettor allocates
-nothing per step, and the `agrapa` and `ons` bettors allocate a small
-constant per-step state record.
-
You should compile with the `llvm` flag for maximum performance.
## Development
diff --git a/ppad-eproc.cabal b/ppad-eproc.cabal
@@ -14,8 +14,7 @@ description:
Anytime-valid sequential hypothesis testing for bounded random
variables, via the e-process / betting framework of Waudby-Smith and
Ramdas (2024). Provides bounded-mean and paired two-sample tests
- with fixed-lambda, aGRAPA, and ONS bettors. Tests are valid under
- optional stopping.
+ with fixed-lambda, aGRAPA, and ONS bettors.
flag llvm
description: Use GHC's LLVM backend.
