valentin/BigData
1
0
Fork 0
mirror of https://github.com/Vale54321/BigData.git synced 2025-12-15 11:29:32 +01:00
Code Issues Packages Projects Releases Wiki Activity

Aufgabe 10

Browse Source
This commit is contained in:
Valentin Heiserer
2025-12-11 20:39:21 +01:00
parent de3782d570
commit f89d39d420
1 changed files with 79 additions and 84 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
Aufgabe 10/Aufgabe10.py
View File

@@ -62,7 +62,7 @@ def compute_daily_and_yearly_frosts(spark: SparkSession):
q_daily_max = (
"SELECT stationId, date, SUBSTR(CAST(date AS STRING),1,4) AS year, MAX(TT_TU) AS max_temp "
"FROM german_stations_data "
"WHERE TT_TU IS NOT NULL "
"WHERE TT_TU IS NOT NULL AND TT_TU > -50 AND TT_TU < 60 "
"GROUP BY stationId, date"
)
daily_max = spark.sql(q_daily_max)
@@ -88,9 +88,6 @@ def compute_daily_and_yearly_frosts(spark: SparkSession):
def frost_analysis(spark: SparkSession, year=2024, station_name_matches=('kempten',)):
compute_daily_and_yearly_frosts(spark)
# Debug: check available years and data
spark.sql("SELECT year, COUNT(*) as cnt FROM station_year_frost GROUP BY year ORDER BY year").show(50)
q_hist = (
f"SELECT frost_days, COUNT(*) AS station_count "
f"FROM station_year_frost WHERE year = '{year}' GROUP BY frost_days ORDER BY frost_days"
@@ -101,7 +98,6 @@ def frost_analysis(spark: SparkSession, year=2024, station_name_matches=('kempte
if hist_pdf.empty:
print(f"No frost data found for year {year}. Trying to find available years...")
# Try without year filter to see if data exists
q_all = "SELECT frost_days, COUNT(*) AS station_count FROM station_year_frost GROUP BY frost_days ORDER BY frost_days"
hist_pdf = spark.sql(q_all).toPandas()
if hist_pdf.empty:
@@ -129,11 +125,10 @@ def frost_analysis(spark: SparkSession, year=2024, station_name_matches=('kempte
sname = r['station_name']
print(f"Analyzing stationId={sid} name={sname}")
# compute frostdays + 5-yr and 20-yr rolling averages using window frame
q_ts = (
"SELECT year, frost_days, "
"AVG(frost_days) OVER (PARTITION BY stationId ORDER BY CAST(year AS INT) ROWS BETWEEN 4 PRECEDING AND CURRENT ROW) AS avg_5, "
"AVG(frost_days) OVER (PARTITION BY stationId ORDER BY CAST(year AS INT) ROWS BETWEEN 19 PRECEDING AND CURRENT ROW) AS avg_20 "
"AVG(frost_days) OVER (PARTITION BY stationId ORDER BY CAST(year AS INT) RANGE BETWEEN 4 PRECEDING AND CURRENT ROW) AS avg_5, "
"AVG(frost_days) OVER (PARTITION BY stationId ORDER BY CAST(year AS INT) RANGE BETWEEN 19 PRECEDING AND CURRENT ROW) AS avg_20 "
f"FROM station_year_frost WHERE stationId = {sid} ORDER BY CAST(year AS INT)"
)
ts_df = spark.sql(q_ts)