feat(user): add dominant topic information to user data
This commit is contained in:
@@ -88,6 +88,15 @@ export default function UserModal({
|
||||
</div>
|
||||
</div>
|
||||
) : null}
|
||||
|
||||
{userData.dominant_topic ? (
|
||||
<div style={styles.topUserItem}>
|
||||
<div style={styles.topUserName}>Most Common Topic</div>
|
||||
<div style={styles.topUserMeta}>
|
||||
{userData.dominant_topic.topic} ({userData.dominant_topic.count} events)
|
||||
</div>
|
||||
</div>
|
||||
) : null}
|
||||
</div>
|
||||
)}
|
||||
</DialogPanel>
|
||||
|
||||
@@ -34,6 +34,11 @@ type Vocab = {
|
||||
top_words: FrequencyWord[];
|
||||
};
|
||||
|
||||
type DominantTopic = {
|
||||
topic: string;
|
||||
count: number;
|
||||
};
|
||||
|
||||
type User = {
|
||||
author: string;
|
||||
post: number;
|
||||
@@ -41,6 +46,7 @@ type User = {
|
||||
comment_post_ratio: number;
|
||||
comment_share: number;
|
||||
avg_emotions?: Record<string, number>;
|
||||
dominant_topic?: DominantTopic | null;
|
||||
vocab?: Vocab | null;
|
||||
};
|
||||
|
||||
@@ -202,6 +208,7 @@ type FilterResponse = {
|
||||
|
||||
export type {
|
||||
TopUser,
|
||||
DominantTopic,
|
||||
Vocab,
|
||||
User,
|
||||
InteractionGraph,
|
||||
|
||||
@@ -71,6 +71,7 @@ class UserAnalysis:
|
||||
per_user = df.groupby(["author", "type"]).size().unstack(fill_value=0)
|
||||
|
||||
emotion_cols = [col for col in df.columns if col.startswith("emotion_")]
|
||||
dominant_topic_by_author = {}
|
||||
|
||||
avg_emotions_by_author = {}
|
||||
if emotion_cols:
|
||||
@@ -80,6 +81,31 @@ class UserAnalysis:
|
||||
for author, row in avg_emotions.iterrows()
|
||||
}
|
||||
|
||||
if "topic" in df.columns:
|
||||
topic_df = df[
|
||||
df["topic"].notna()
|
||||
& (df["topic"] != "")
|
||||
& (df["topic"] != "Misc")
|
||||
]
|
||||
if not topic_df.empty:
|
||||
topic_counts = (
|
||||
topic_df.groupby(["author", "topic"])
|
||||
.size()
|
||||
.reset_index(name="count")
|
||||
.sort_values(
|
||||
["author", "count", "topic"],
|
||||
ascending=[True, False, True],
|
||||
)
|
||||
.drop_duplicates(subset=["author"])
|
||||
)
|
||||
dominant_topic_by_author = {
|
||||
row["author"]: {
|
||||
"topic": row["topic"],
|
||||
"count": int(row["count"]),
|
||||
}
|
||||
for _, row in topic_counts.iterrows()
|
||||
}
|
||||
|
||||
# ensure columns always exist
|
||||
for col in ("post", "comment"):
|
||||
if col not in per_user.columns:
|
||||
@@ -109,6 +135,7 @@ class UserAnalysis:
|
||||
"comment_post_ratio": float(row.get("comment_post_ratio", 0)),
|
||||
"comment_share": float(row.get("comment_share", 0)),
|
||||
"avg_emotions": avg_emotions_by_author.get(author, {}),
|
||||
"dominant_topic": dominant_topic_by_author.get(author),
|
||||
"vocab": vocab_by_author.get(
|
||||
author,
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user