edfcore 0.4.314 → 0.4.316

package/dist/constants.d.ts

@@ -109,5 +109,5 @@ export declare const SIGNAL_FIELD_BLOCK_OFFSETS: {
     readonly reserved: 224;
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export declare const VERSION = "0.4.314";
+export declare const VERSION = "0.4.316";
 //# sourceMappingURL=constants.d.ts.map

package/dist/constants.js

@@ -79,5 +79,5 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
     reserved: 224,
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.314';
+export const VERSION = '0.4.316';
 //# sourceMappingURL=constants.js.map

package/docs/CHANGELOG.md

@@ -6,6 +6,29 @@ alone does not tell you whether you were affected.
 edfcore is pre-1.0. Patch releases have carried behaviour changes where the old behaviour was a
 defect; those are called out below.
 
+## 0.4.316
+
+- **Added** an execution of the negative-gain section of `physical-values.md`: the scale it prints,
+  the three samples it converts, the envelope `physicalRangeOf` reports in size order rather than
+  field order, and the whole `INVERTED_PHYSICAL_RANGE` message it quotes for a one-signal file,
+  compared word for word against the one the package emits.
+- Only the page's hard wraps are undone for that comparison. A run of spaces inside a line is not
+  wrapping — it is the eight-byte physical minimum field quoted as the file holds it, padding
+  included — so collapsing every space would have compared against a message edfcore does not
+  emit. The byte offset in the quote is resolved through `signalFieldOffset`, which is the same
+  `256 + ns*104 + i*8` the address table on `edf-format.md` gives.
+
+## 0.4.315
+
+- **Added** a measurement of the float32 cost `physical-values.md` cites as the reason `toPhysical`
+  has no `Float32` option. Every one of the 2^24 BDF samples on a -500..500 uV channel is converted
+  through the scale edfcore publishes and rounded to float32, and the worst error is compared with
+  the 0.26 of a quantisation step the page prints.
+- The sentence the number supports is checked too: float32 carries 24 significand bits and a BDF
+  sample is a 24-bit integer, so the digital values themselves survive the round trip exactly and
+  there is nothing left for the scaling. Run as a scalar loop — 2^24 float64 samples is 134 MB, and
+  the point is the worst case, not the array.
+
 ## 0.4.314
 
 - **Added** the census underneath the table 0.4.313 pinned: the two conversion forms are run over

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.314",
+  "version": "0.4.316",
   "description": "Modern, typed, zero-dependency reader for EDF, EDF+, BDF and BDF+ biosignal files. Works in browsers and Node with true random access.",
   "keywords": [
     "edf",

package/src/constants.ts

@@ -93,4 +93,4 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
 } as const;
 
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.314';
+export const VERSION = '0.4.316';