tpattern 0.1.14__tar.gz → 0.1.15__tar.gz

{tpattern-0.1.14/tpattern.egg-info → tpattern-0.1.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tpattern
-Version: 0.1.14
+Version: 0.1.15
 Summary: Open, auditable temporal & sequential pattern analysis (T-patterns) for sport and behaviour
 Author-email: Andrew Callaway <acallaway@bournemouth.ac.uk>
 License-Expression: MIT
@@ -49,7 +49,7 @@ auditable code — the same critical-interval method the THEME software applies
 [The method it implements](#the-method-it-implements)) — and is the analysis
 companion to the [OpenTag.Studio](https://opentag.studio) tagging tool.
 
-## Two ways to use it
+## Three ways to use it
 
 ### 🚀 In your browser — Google Colab, no install, no coding
 
@@ -71,6 +71,15 @@ Three notebooks (all in [`colab/`](colab/)):
 
 `pip install tpattern` and script it — see [Getting started](#getting-started-in-a-minute), [Quickstart](#quickstart) and [`examples/`](examples/).
 
+### 📊 From R
+
+R users (common in psychology and behavioural science) can drive the **same
+engine** via [`reticulate`](https://rstudio.github.io/reticulate/) — the algorithm
+stays in Python, so results match exactly. Install with
+`remotes::install_github("ajcallaway/TPattern", subdir = "r")`, then run the whole
+analysis in one call: `tpattern_analysis("events.csv")`. See
+[Using tpattern from R](#using-tpattern-from-r).
+
 Bring your own data (one row per event — see [`SCHEMA.md`](SCHEMA.md)) or try the
 included [`data/`](data/) example.
 

{tpattern-0.1.14 → tpattern-0.1.15}/README.md

@@ -14,7 +14,7 @@ auditable code — the same critical-interval method the THEME software applies
 [The method it implements](#the-method-it-implements)) — and is the analysis
 companion to the [OpenTag.Studio](https://opentag.studio) tagging tool.
 
-## Two ways to use it
+## Three ways to use it
 
 ### 🚀 In your browser — Google Colab, no install, no coding
 
@@ -36,6 +36,15 @@ Three notebooks (all in [`colab/`](colab/)):
 
 `pip install tpattern` and script it — see [Getting started](#getting-started-in-a-minute), [Quickstart](#quickstart) and [`examples/`](examples/).
 
+### 📊 From R
+
+R users (common in psychology and behavioural science) can drive the **same
+engine** via [`reticulate`](https://rstudio.github.io/reticulate/) — the algorithm
+stays in Python, so results match exactly. Install with
+`remotes::install_github("ajcallaway/TPattern", subdir = "r")`, then run the whole
+analysis in one call: `tpattern_analysis("events.csv")`. See
+[Using tpattern from R](#using-tpattern-from-r).
+
 Bring your own data (one row per event — see [`SCHEMA.md`](SCHEMA.md)) or try the
 included [`data/`](data/) example.
 

{tpattern-0.1.14 → tpattern-0.1.15}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "tpattern"
-version = "0.1.14"
+version = "0.1.15"
 description = "Open, auditable temporal & sequential pattern analysis (T-patterns) for sport and behaviour"
 readme = "README.md"
 requires-python = ">=3.9"

{tpattern-0.1.14 → tpattern-0.1.15}/tpattern/__init__.py

@@ -6,7 +6,7 @@ behavior: T-patterns and their detection. Behavior Research Methods,
 Instruments, & Computers, 32(1), 93-110.
 """
 
-__version__ = "0.1.14"   # defined before submodule imports (methods.py reads it)
+__version__ = "0.1.15"   # defined before submodule imports (methods.py reads it)
 
 from .io import (Observation, read_observation, read_sample, read_table,
                  read_sofcoder, DataFormatError, check_reference, CHECKS)

{tpattern-0.1.14 → tpattern-0.1.15}/tpattern/significance.py

@@ -66,6 +66,7 @@ class Calibrated:
     p_emp_strength: float = 1.0   # empirical p of the analytic strength
     fwer_keep: bool = False   # Holm on p_emp
     fdr_q: float = 1.0        # Benjamini-Hochberg q on p_emp
+    fdr_q_by: float = 1.0     # Benjamini-Yekutieli q (dependence-robust FDR)
 
 
 @dataclass
@@ -81,6 +82,8 @@ class CalibrationResult:
     def kept(self, method: str = "fdr") -> list[Calibrated]:
         if method == "fwer":
             return [c for c in self.real if c.fwer_keep]
+        if method in ("fdr_by", "by"):
+            return [c for c in self.real if c.fdr_q_by <= self.q_target]
         return [c for c in self.real if c.fdr_q <= self.q_target]
 
 
@@ -100,6 +103,20 @@ def _benjamini_hochberg(items, get_p):
     return {id(items[i]): q[i] for i in range(m)}
 
 
+def _benjamini_yekutieli(items, get_p):
+    """Benjamini-Yekutieli q-values: the BH q scaled by the harmonic factor
+    c(m) = sum_{i=1..m} 1/i, which gives false-discovery-rate control under
+    arbitrary dependence among the p-values (Benjamini & Yekutieli, 2001).
+    Surrogate p-values for patterns that share events are dependent, so this is
+    the dependence-robust companion to the BH q; it is more conservative."""
+    m = len(items)
+    if m == 0:
+        return {}
+    cm = sum(1.0 / i for i in range(1, m + 1))
+    bh = _benjamini_hochberg(items, get_p)
+    return {k: min(1.0, v * cm) for k, v in bh.items()}
+
+
 def _holm_keep(items, get_p, alpha):
     """Holm–Bonferroni step-down; return set of ids kept at family alpha."""
     m = len(items)
@@ -217,9 +234,11 @@ def calibrate(observations, config: Config | None = None, *,
         grp = [c for c in real if c.level == lv]
         keep = _holm_keep(grp, lambda c: c.p_emp, alpha)
         qmap = _benjamini_hochberg(grp, lambda c: c.p_emp)
+        bymap = _benjamini_yekutieli(grp, lambda c: c.p_emp)
         for c in grp:
             c.fwer_keep = id(c) in keep
             c.fdr_q = qmap[id(c)]
+            c.fdr_q_by = bymap[id(c)]
 
     return CalibrationResult(real=real, surrogate_counts=surrogate_counts,
                              B=B, alpha=alpha, q_target=q_target, null=null,

{tpattern-0.1.14 → tpattern-0.1.15/tpattern.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tpattern
-Version: 0.1.14
+Version: 0.1.15
 Summary: Open, auditable temporal & sequential pattern analysis (T-patterns) for sport and behaviour
 Author-email: Andrew Callaway <acallaway@bournemouth.ac.uk>
 License-Expression: MIT
@@ -49,7 +49,7 @@ auditable code — the same critical-interval method the THEME software applies
 [The method it implements](#the-method-it-implements)) — and is the analysis
 companion to the [OpenTag.Studio](https://opentag.studio) tagging tool.
 
-## Two ways to use it
+## Three ways to use it
 
 ### 🚀 In your browser — Google Colab, no install, no coding
 
@@ -71,6 +71,15 @@ Three notebooks (all in [`colab/`](colab/)):
 
 `pip install tpattern` and script it — see [Getting started](#getting-started-in-a-minute), [Quickstart](#quickstart) and [`examples/`](examples/).
 
+### 📊 From R
+
+R users (common in psychology and behavioural science) can drive the **same
+engine** via [`reticulate`](https://rstudio.github.io/reticulate/) — the algorithm
+stays in Python, so results match exactly. Install with
+`remotes::install_github("ajcallaway/TPattern", subdir = "r")`, then run the whole
+analysis in one call: `tpattern_analysis("events.csv")`. See
+[Using tpattern from R](#using-tpattern-from-r).
+
 Bring your own data (one row per event — see [`SCHEMA.md`](SCHEMA.md)) or try the
 included [`data/`](data/) example.