Metrix/InterferenceETL
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: 邻近高干扰小区数按制式拆分统计

Browse Source
This commit is contained in:
nixevol
2026-08-11 16:59:48 +08:00
parent eca4c5cbb1
commit 3835a0b928
4 changed files with 126 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files
main.py
+57 -16
View File
@@ -72,6 +72,13 @@ MIN_INTERFERENCE_BY_SOURCE = {
EARTH_RADIUS_KM = 6371.0088
NEARBY_RADIUS_KM = 1.0
NEARBY_COLUMN_BY_NETWORK_TYPE = {
"2.6G": "nearby_26g",
"700M": "nearby_700m",
"TDD": "nearby_tdd",
"FDD": "nearby_fdd",
}
DEFAULT_STORAGE_ID = "stg_4d9a910d72"
DEFAULT_ROOT = "/网优日常优化数据文档/(勿删)干扰定时小时指标"
CELL_DATA_DIRECTORIES = (
@@ -192,7 +199,15 @@ SUMMARY_HEADER = (
"latitude",
"azimuth",
"nearby_count",
"nearby_26g",
"nearby_700m",
"nearby_tdd",
"nearby_fdd",
"prev_nearby_count",
"prev_nearby_26g",
"prev_nearby_700m",
"prev_nearby_tdd",
"prev_nearby_fdd",
)
class ProcessingError(RuntimeError):
@@ -358,7 +373,7 @@ class ApiSummaryStore:
START TRANSACTION;
DELETE FROM `{self.table}` WHERE metric_time = {metric_literal};
INSERT INTO `{self.table}`
(metric_time, network_type, cgi, cell_name, interference_dbm, prev_interference_dbm, longitude, latitude, azimuth, nearby_count, prev_nearby_count)
(metric_time, network_type, cgi, cell_name, interference_dbm, prev_interference_dbm, longitude, latitude, azimuth, nearby_count, nearby_26g, nearby_700m, nearby_tdd, nearby_fdd, prev_nearby_count, prev_nearby_26g, prev_nearby_700m, prev_nearby_tdd, prev_nearby_fdd)
VALUES
{values};
DELETE FROM `{self.table}` WHERE metric_time <> {metric_literal};
@@ -392,8 +407,7 @@ class ApiSummaryStore:
self._ensure_schema()
metric_literal = f"'{metric_time:%Y-%m-%d %H:%M:%S}'"
sql = (
"SELECT metric_time, network_type, cgi, cell_name, interference_dbm, "
"prev_interference_dbm, longitude, latitude, azimuth, nearby_count, prev_nearby_count "
f"SELECT {', '.join(SUMMARY_HEADER)} "
f"FROM `{self.table}` "
f"WHERE metric_time = {metric_literal} ORDER BY cgi, cell_name"
)
@@ -431,7 +445,15 @@ class ApiSummaryStore:
latitude DECIMAL(10,6) NULL,
azimuth DECIMAL(6,2) NOT NULL DEFAULT 0,
nearby_count INT NOT NULL DEFAULT 0,
nearby_26g INT NOT NULL DEFAULT 0,
nearby_700m INT NOT NULL DEFAULT 0,
nearby_tdd INT NOT NULL DEFAULT 0,
nearby_fdd INT NOT NULL DEFAULT 0,
prev_nearby_count INT NULL,
prev_nearby_26g INT NULL,
prev_nearby_700m INT NULL,
prev_nearby_tdd INT NULL,
prev_nearby_fdd INT NULL,
PRIMARY KEY (metric_time, cgi)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4
""",
@@ -451,7 +473,15 @@ class ApiSummaryStore:
"latitude": "DECIMAL(10,6) NULL",
"azimuth": "DECIMAL(6,2) NOT NULL DEFAULT 0",
"nearby_count": "INT NOT NULL DEFAULT 0",
"nearby_26g": "INT NOT NULL DEFAULT 0",
"nearby_700m": "INT NOT NULL DEFAULT 0",
"nearby_tdd": "INT NOT NULL DEFAULT 0",
"nearby_fdd": "INT NOT NULL DEFAULT 0",
"prev_nearby_count": "INT NULL",
"prev_nearby_26g": "INT NULL",
"prev_nearby_700m": "INT NULL",
"prev_nearby_tdd": "INT NULL",
"prev_nearby_fdd": "INT NULL",
}
for column, definition in migrations.items():
if column not in existing_columns:
@@ -487,7 +517,15 @@ class ApiSummaryStore:
sql_decimal_literal(row["latitude"], "latitude", nullable=True),
sql_decimal_literal(row["azimuth"], "azimuth"),
str(int(row["nearby_count"])),
str(int(row["nearby_26g"])),
str(int(row["nearby_700m"])),
str(int(row["nearby_tdd"])),
str(int(row["nearby_fdd"])),
sql_decimal_literal(row["prev_nearby_count"], "prev_nearby_count", nullable=True),
sql_decimal_literal(row["prev_nearby_26g"], "prev_nearby_26g", nullable=True),
sql_decimal_literal(row["prev_nearby_700m"], "prev_nearby_700m", nullable=True),
sql_decimal_literal(row["prev_nearby_tdd"], "prev_nearby_tdd", nullable=True),
sql_decimal_literal(row["prev_nearby_fdd"], "prev_nearby_fdd", nullable=True),
)
) + ")"
@@ -966,7 +1004,7 @@ def summary_record(
interference = required(record, "载波平均噪声干扰(dBm)", source_path, row_number)
longitude, latitude, azimuth = (cell_metadata or {}).get(cgi, ("", "", "0"))
return {
record_out = {
"metric_time": metric_time.strftime("%Y-%m-%d %H:%M:%S"),
"network_type": NETWORK_TYPE_BY_SOURCE[source_type],
"cgi": cgi,
@@ -979,6 +1017,10 @@ def summary_record(
"nearby_count": "0",
"prev_nearby_count": "",
}
for column in NEARBY_COLUMN_BY_NETWORK_TYPE.values():
record_out[column] = "0"
record_out[f"prev_{column}"] = ""
return record_out
def passes_interference_threshold(record: dict[str, str], source_type: str) -> bool:
@@ -992,10 +1034,9 @@ def passes_interference_threshold(record: dict[str, str], source_type: str) -> b
def populate_nearby_counts(rows: list[dict[str, str]]) -> None:
counts = [0] * len(rows)
type_counts = [dict.fromkeys(NEARBY_COLUMN_BY_NETWORK_TYPE.values(), 0) for _ in rows]
spatial_index: dict[tuple[int, int, int], list[tuple[int, float, float]]] = {}
for index, row in enumerate(rows):
row["nearby_count"] = "0"
if not row["longitude"] or not row["latitude"]:
continue
try:
@@ -1006,6 +1047,7 @@ def populate_nearby_counts(rows: list[dict[str, str]]) -> None:
if not math.isfinite(longitude) or not math.isfinite(latitude) or not -180 <= longitude <= 180 or not -90 <= latitude <= 90:
raise ProcessingError(f"Invalid coordinates for CGI {row['cgi']}")
column = NEARBY_COLUMN_BY_NETWORK_TYPE[row["network_type"]]
bucket = spatial_bucket(longitude, latitude)
for x_offset in (-1, 0, 1):
for y_offset in (-1, 0, 1):
@@ -1013,12 +1055,14 @@ def populate_nearby_counts(rows: list[dict[str, str]]) -> None:
nearby_bucket = (bucket[0] + x_offset, bucket[1] + y_offset, bucket[2] + z_offset)
for other_index, other_longitude, other_latitude in spatial_index.get(nearby_bucket, ()):
if haversine_distance_km(longitude, latitude, other_longitude, other_latitude) <= NEARBY_RADIUS_KM:
counts[index] += 1
counts[other_index] += 1
type_counts[index][NEARBY_COLUMN_BY_NETWORK_TYPE[rows[other_index]["network_type"]]] += 1
type_counts[other_index][column] += 1
spatial_index.setdefault(bucket, []).append((index, longitude, latitude))
for index, count in enumerate(counts):
rows[index]["nearby_count"] = str(count)
for row, counts in zip(rows, type_counts, strict=True):
for column, count in counts.items():
row[column] = str(count)
row["nearby_count"] = str(sum(counts.values()))
def deduplicate_by_cgi(rows: list[dict[str, str]]) -> int:
@@ -1038,14 +1082,11 @@ def deduplicate_by_cgi(rows: list[dict[str, str]]) -> int:
def populate_previous_values(rows: list[dict[str, str]], previous_rows: list[dict[str, str]]) -> None:
previous_by_cgi = {row["cgi"]: row for row in previous_rows}
copied_columns = ("interference_dbm", "nearby_count", *NEARBY_COLUMN_BY_NETWORK_TYPE.values())
for row in rows:
previous = previous_by_cgi.get(row["cgi"])
if previous is None:
row["prev_interference_dbm"] = ""
row["prev_nearby_count"] = ""
continue
row["prev_interference_dbm"] = previous["interference_dbm"]
row["prev_nearby_count"] = previous["nearby_count"]
for column in copied_columns:
row[f"prev_{column}"] = previous[column] if previous is not None else ""
def spatial_bucket(longitude: float, latitude: float) -> tuple[int, int, int]: