43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363 | class IdentifyMissingDataPattern(MonitoringBaseInterface, InputValidator):
"""
Identifies missing data in a DataFrame based on specified time patterns.
Logs the expected missing times.
Args:
df (pyspark.sql.Dataframe): DataFrame containing at least the 'EventTime' column.
patterns (list of dict): List of dictionaries specifying the time patterns.
- For 'minutely' frequency: Specify 'second' and optionally 'millisecond'.
Example: [{'second': 0}, {'second': 13}, {'second': 49}]
- For 'hourly' frequency: Specify 'minute', 'second', and optionally 'millisecond'.
Example: [{'minute': 0, 'second': 0}, {'minute': 30, 'second': 30}]
frequency (str): Frequency of the patterns. Must be either 'minutely' or 'hourly'.
- 'minutely': Patterns are checked every minute at specified seconds.
- 'hourly': Patterns are checked every hour at specified minutes and seconds.
tolerance (str, optional): Maximum allowed deviation from the pattern (e.g., '1s', '500ms').
Default is '10ms'.
Example:
```python
from pyspark.sql import SparkSession
spark = SparkSession.builder.master("local[1]").appName("IdentifyMissingDataPatternExample").getOrCreate()
patterns = [
{"second": 0},
{"second": 20},
]
frequency = "minutely"
tolerance = "1s"
identify_missing_data = IdentifyMissingDataPattern(
df=df,
patterns=patterns,
frequency=frequency,
tolerance=tolerance,
)
identify_missing_data.check()
```
"""
df: PySparkDataFrame
EXPECTED_SCHEMA = StructType(
[
StructField("TagName", StringType(), True),
StructField("EventTime", TimestampType(), True),
StructField("Status", StringType(), True),
StructField("Value", FloatType(), True),
]
)
def __init__(
self,
df: PySparkDataFrame,
patterns: list,
frequency: str = "minutely",
tolerance: str = "10ms",
) -> None:
self.df = df
self.patterns = patterns
self.frequency = frequency.lower()
self.tolerance = tolerance
self.validate(self.EXPECTED_SCHEMA)
# Configure logging
self.logger = LoggerManager().create_logger(self.__class__.__name__)
@staticmethod
def system_type():
"""
Attributes:
SystemType (Environment): Requires PYSPARK
"""
return SystemType.PYSPARK
@staticmethod
def libraries():
libraries = Libraries()
return libraries
@staticmethod
def settings() -> dict:
return {}
def check(self) -> PySparkDataFrame:
"""
Executes the missing pattern detection logic. Identifies and logs any missing patterns
based on the provided patterns and frequency within the specified tolerance.
Returns:
pyspark.sql.DataFrame:
Returns the original PySpark DataFrame without changes.
"""
self._validate_inputs()
df = self.df.withColumn("EventTime", F.to_timestamp("EventTime"))
df_sorted = df.orderBy("EventTime")
# Determine if the DataFrame is empty
count = df_sorted.count()
if count == 0:
self.logger.info("Generated 0 expected times based on patterns.")
self.logger.info("DataFrame is empty. No missing patterns to detect.")
return self.df
# Determine the time range of the data
min_time, max_time = df_sorted.agg(
F.min("EventTime"), F.max("EventTime")
).first()
if not min_time or not max_time:
self.logger.info("Generated 0 expected times based on patterns.")
self.logger.info("DataFrame is empty. No missing patterns to detect.")
return self.df
# Generate all expected times based on patterns and frequency
expected_times_df = self._generate_expected_times(min_time, max_time)
# Identify missing patterns by left joining expected times with actual EventTimes within tolerance
missing_patterns_df = self._find_missing_patterns(expected_times_df, df_sorted)
self._log_missing_patterns(missing_patterns_df)
return self.df
def _validate_inputs(self):
if self.frequency not in ["minutely", "hourly"]:
error_msg = "Frequency must be either 'minutely' or 'hourly'."
self.logger.error(error_msg)
raise ValueError(error_msg)
for pattern in self.patterns:
if self.frequency == "minutely":
if "second" not in pattern:
raise ValueError(
"Each pattern must have a 'second' key for 'minutely' frequency."
)
if pattern.get("second", 0) >= 60:
raise ValueError(
"For 'minutely' frequency, 'second' must be less than 60."
)
if "minute" in pattern or "hour" in pattern:
raise ValueError(
"For 'minutely' frequency, pattern should not contain 'minute' or 'hour'."
)
elif self.frequency == "hourly":
if "minute" not in pattern or "second" not in pattern:
raise ValueError(
"Each pattern must have 'minute' and 'second' keys for 'hourly' frequency."
)
if pattern.get("minute", 0) >= 60:
raise ValueError(
"For 'hourly' frequency, 'minute' must be less than 60."
)
if "hour" in pattern:
raise ValueError(
"For 'hourly' frequency, pattern should not contain 'hour'."
)
try:
self.tolerance_ms = parse_time_string_to_ms(self.tolerance)
self.tolerance_seconds = self.tolerance_ms / 1000
self.logger.info(
f"Using tolerance: {self.tolerance_ms} ms ({self.tolerance_seconds} seconds)"
)
except ValueError as e:
error_msg = f"Invalid tolerance format: {self.tolerance}"
self.logger.error(error_msg)
raise ValueError(error_msg) from e
def _generate_expected_times(self, min_time, max_time) -> PySparkDataFrame:
floor_min_time = self._get_floor_min_time(min_time)
ceil_max_time = self._get_ceil_max_time(max_time)
base_times_df = self._create_base_times_df(floor_min_time, ceil_max_time)
expected_times_df = self._apply_patterns(
base_times_df, floor_min_time, max_time
)
return expected_times_df
def _get_floor_min_time(self, min_time):
if self.frequency == "minutely":
return min_time.replace(second=0, microsecond=0)
elif self.frequency == "hourly":
return min_time.replace(minute=0, second=0, microsecond=0)
def _get_ceil_max_time(self, max_time):
if self.frequency == "minutely":
return (max_time + pd.Timedelta(minutes=1)).replace(second=0, microsecond=0)
elif self.frequency == "hourly":
return (max_time + pd.Timedelta(hours=1)).replace(
minute=0, second=0, microsecond=0
)
def _create_base_times_df(self, floor_min_time, ceil_max_time):
step = F.expr(f"INTERVAL 1 {self.frequency.upper()[:-2]}")
return self.df.sparkSession.createDataFrame(
[(floor_min_time, ceil_max_time)], ["start", "end"]
).select(
F.explode(
F.sequence(
F.col("start").cast("timestamp"),
F.col("end").cast("timestamp"),
step,
)
).alias("BaseTime")
)
def _apply_patterns(self, base_times_df, floor_min_time, max_time):
expected_times = []
for pattern in self.patterns:
expected_time = self._calculate_expected_time(base_times_df, pattern)
expected_times.append(expected_time)
expected_times_df = (
base_times_df.withColumn(
"ExpectedTime", F.explode(F.array(*expected_times))
)
.select("ExpectedTime")
.distinct()
.filter(
(F.col("ExpectedTime") >= F.lit(floor_min_time))
& (F.col("ExpectedTime") <= F.lit(max_time))
)
)
return expected_times_df
def _calculate_expected_time(self, base_times_df, pattern):
if self.frequency == "minutely":
seconds = pattern.get("second", 0)
milliseconds = pattern.get("millisecond", 0)
return (
F.col("BaseTime")
+ F.expr(f"INTERVAL {seconds} SECOND")
+ F.expr(f"INTERVAL {milliseconds} MILLISECOND")
)
elif self.frequency == "hourly":
minutes = pattern.get("minute", 0)
seconds = pattern.get("second", 0)
milliseconds = pattern.get("millisecond", 0)
return (
F.col("BaseTime")
+ F.expr(f"INTERVAL {minutes} MINUTE")
+ F.expr(f"INTERVAL {seconds} SECOND")
+ F.expr(f"INTERVAL {milliseconds} MILLISECOND")
)
def _find_missing_patterns(
self, expected_times_df: PySparkDataFrame, actual_df: PySparkDataFrame
) -> PySparkDataFrame:
"""
Finds missing patterns by comparing expected times with actual EventTimes within tolerance.
Args:
expected_times_df (PySparkDataFrame): DataFrame with expected 'ExpectedTime'.
actual_df (PySparkDataFrame): Actual DataFrame with 'EventTime'.
Returns:
PySparkDataFrame: DataFrame with missing 'ExpectedTime'.
"""
# Format tolerance for SQL INTERVAL
tolerance_str = self._format_timedelta_for_sql(self.tolerance_ms)
# Perform left join with tolerance window
missing_patterns_df = (
expected_times_df.alias("et")
.join(
actual_df.alias("at"),
(
F.col("at.EventTime")
>= F.expr(f"et.ExpectedTime - INTERVAL {tolerance_str}")
)
& (
F.col("at.EventTime")
<= F.expr(f"et.ExpectedTime + INTERVAL {tolerance_str}")
),
how="left",
)
.filter(F.col("at.EventTime").isNull())
.select(F.col("et.ExpectedTime"))
)
self.logger.info(f"Identified {missing_patterns_df.count()} missing patterns.")
return missing_patterns_df
def _log_missing_patterns(self, missing_patterns_df: PySparkDataFrame):
"""
Logs the missing patterns.
Args:
missing_patterns_df (PySparkDataFrame): DataFrame with missing 'ExpectedTime'.
"""
missing_patterns = missing_patterns_df.collect()
if missing_patterns:
self.logger.info("Detected Missing Patterns:")
# Sort missing patterns by ExpectedTime
sorted_missing_patterns = sorted(
missing_patterns, key=lambda row: row["ExpectedTime"]
)
for row in sorted_missing_patterns:
# Format ExpectedTime to include milliseconds correctly
formatted_time = row["ExpectedTime"].strftime("%Y-%m-%d %H:%M:%S.%f")[
:-3
]
self.logger.info(f"Missing Pattern at {formatted_time}")
else:
self.logger.info("No missing patterns detected.")
@staticmethod
def _format_timedelta_for_sql(tolerance_ms: float) -> str:
"""
Formats a tolerance in milliseconds to a string suitable for SQL INTERVAL.
Args:
tolerance_ms (float): Tolerance in milliseconds.
Returns:
str: Formatted string (e.g., '1 SECOND', '500 MILLISECONDS').
"""
if tolerance_ms >= 3600000:
hours = int(tolerance_ms // 3600000)
return f"{hours} HOURS"
elif tolerance_ms >= 60000:
minutes = int(tolerance_ms // 60000)
return f"{minutes} MINUTES"
elif tolerance_ms >= 1000:
seconds = int(tolerance_ms // 1000)
return f"{seconds} SECONDS"
else:
milliseconds = int(tolerance_ms)
return f"{milliseconds} MILLISECONDS"
|