From 988217b34cae7d1ca8681c93165e50084034fadc Mon Sep 17 00:00:00 2001
From: William FH <13333726+hinthornw@users.noreply.github.com>
Date: Tue, 3 Sep 2024 09:01:46 -0700
Subject: [PATCH] Filter child runs where trace has been filtered (#964)

---
 js/package.json            |  2 +-
 js/src/client.ts           | 21 +++++++++++----------
 js/src/index.ts            |  2 +-
 python/langsmith/client.py | 19 ++++++++++++-------
 python/pyproject.toml      |  2 +-
 5 files changed, 26 insertions(+), 20 deletions(-)

diff --git a/js/package.json b/js/package.json
index e9681e5ae..c3ff5e32d 100644
--- a/js/package.json
+++ b/js/package.json
@@ -1,6 +1,6 @@
 {
   "name": "langsmith",
-  "version": "0.1.49",
+  "version": "0.1.50",
   "description": "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform.",
   "packageManager": "yarn@1.22.19",
   "files": [
diff --git a/js/src/client.ts b/js/src/client.ts
index 484cd91d5..4d7538bda 100644
--- a/js/src/client.ts
+++ b/js/src/client.ts
@@ -400,7 +400,7 @@ export class Client {
 
   private tracingSampleRate?: number;
 
-  private sampledPostUuids = new Set();
+  private filteredPostUuids = new Set();
 
   private autoBatchTracing = true;
 
@@ -648,23 +648,24 @@ export class Client {
     if (patch) {
       const sampled = [];
       for (const run of runs) {
-        if (this.sampledPostUuids.has(run.id)) {
+        if (!this.filteredPostUuids.has(run.id)) {
           sampled.push(run);
-          this.sampledPostUuids.delete(run.id);
+        } else {
+          this.filteredPostUuids.delete(run.id);
         }
       }
       return sampled;
     } else {
       const sampled = [];
       for (const run of runs) {
-        if (run.id !== run.trace_id) {
-          sampled.push(run);
-          this.sampledPostUuids.add(run.id);
-          continue;
-        }
-        if (Math.random() < this.tracingSampleRate) {
+        if (
+          (run.id !== run.trace_id &&
+            !this.filteredPostUuids.has(run.trace_id)) ||
+          Math.random() < this.tracingSampleRate
+        ) {
           sampled.push(run);
-          this.sampledPostUuids.add(run.id);
+        } else {
+          this.filteredPostUuids.add(run.id);
         }
       }
       return sampled;
diff --git a/js/src/index.ts b/js/src/index.ts
index 6e930d3df..bccaa0015 100644
--- a/js/src/index.ts
+++ b/js/src/index.ts
@@ -12,4 +12,4 @@ export type {
 export { RunTree, type RunTreeConfig } from "./run_trees.js";
 
 // Update using yarn bump-version
-export const __version__ = "0.1.49";
+export const __version__ = "0.1.50";
diff --git a/python/langsmith/client.py b/python/langsmith/client.py
index 077a1cfca..4c90ced1c 100644
--- a/python/langsmith/client.py
+++ b/python/langsmith/client.py
@@ -439,7 +439,7 @@ class Client:
         "_web_url",
         "_tenant_id",
         "tracing_sample_rate",
-        "_sampled_post_uuids",
+        "_filtered_post_uuids",
         "tracing_queue",
         "_anonymizer",
         "_hide_inputs",
@@ -524,7 +524,7 @@ def __init__(
             )
 
         self.tracing_sample_rate = _get_tracing_sampling_rate()
-        self._sampled_post_uuids: set[uuid.UUID] = set()
+        self._filtered_post_uuids: set[uuid.UUID] = set()
         self._write_api_urls: Mapping[str, Optional[str]] = _get_write_api_urls(
             api_urls
         )
@@ -1175,19 +1175,24 @@ def _filter_for_sampling(
             sampled = []
             for run in runs:
                 run_id = _as_uuid(run["id"])
-                if run_id in self._sampled_post_uuids:
+                if run_id not in self._filtered_post_uuids:
                     sampled.append(run)
-                    self._sampled_post_uuids.remove(run_id)
+                else:
+                    self._filtered_post_uuids.remove(run_id)
             return sampled
         else:
             sampled = []
             for run in runs:
                 if (
+                    # Child run
                     run["id"] != run.get("trace_id")
-                    or random.random() < self.tracing_sample_rate
-                ):
+                    # Whose trace is included
+                    and run.get("trace_id") not in self._filtered_post_uuids
+                    # Or a root that's randomly sampled
+                ) or random.random() < self.tracing_sample_rate:
                     sampled.append(run)
-                    self._sampled_post_uuids.add(_as_uuid(run["id"]))
+                else:
+                    self._filtered_post_uuids.add(_as_uuid(run["id"]))
             return sampled
 
     def create_run(
diff --git a/python/pyproject.toml b/python/pyproject.toml
index 9b7dd9e26..1aeb0c944 100644
--- a/python/pyproject.toml
+++ b/python/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langsmith"
-version = "0.1.108"
+version = "0.1.109"
 description = "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform."
 authors = ["LangChain <support@langchain.dev>"]
 license = "MIT"