koodiklinikka/palkkakysely
1
0
Fork 0
mirror of https://github.com/koodiklinikka/palkkakysely.git synced 2026-03-12 20:03:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Apply some data fixes

Browse Source
This commit is contained in:
Aarni Koskela
2022-10-10 12:07:48 +03:00
parent 938e37c720
commit 335cf15064
2 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
pulkka/copy_massaged_data.py
View File

@@ -4,7 +4,7 @@ from pulkka.data_ingest import read_data
def main(): def main():
df = read_data() df = read_data()
df.to_html(OUT_DIR / "data.html", index=False) df.to_html(OUT_DIR / "data.html", index=False, na_rep="")
df.to_csv(OUT_DIR / "data.csv", index=False) df.to_csv(OUT_DIR / "data.csv", index=False)
df.to_excel(OUT_DIR / "data.xlsx", index=False) df.to_excel(OUT_DIR / "data.xlsx", index=False)
df.to_json( df.to_json(

4
pulkka/data_ingest.py
View File

@@ -106,8 +106,10 @@ def read_data() -> pd.DataFrame:
# Fill in Vuositulot as 12.5 * Kk-tulot if empty # Fill in Vuositulot as 12.5 * Kk-tulot if empty
df["Vuositulot"] = df.apply(map_vuositulot, axis=1) df["Vuositulot"] = df.apply(map_vuositulot, axis=1)
# Fudge a single known outlier
# Fudge some known outliers
df.loc[df.Vuositulot == 912500, 'Vuositulot'] = 91250 df.loc[df.Vuositulot == 912500, 'Vuositulot'] = 91250
df.loc[df.Kuukausipalkka == 87000, 'Kuukausipalkka'] = 7250
# Synthesize kk-tulot from Vuositulot # Synthesize kk-tulot from Vuositulot
df["Kk-tulot"] = pd.to_numeric(df["Vuositulot"], errors="coerce") / 12 df["Kk-tulot"] = pd.to_numeric(df["Vuositulot"], errors="coerce") / 12