agaric/parse-timelogs-for-upload
10
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Use .apply rather than np.where for easier/better/not-randomly-failing individual parsing

Browse source 
+# On some systems, using np.where worked but others failed.  Why it worked is
+# unknown but why it failed is because numpy where evaluates all parts, even
+# the parts that will never get used because the where clause does not apply!
+# This caused the chained strings to fail because— no string.

This worked fine on the System76 and didn't on Bridget's computer, even after
updating the version of python, but anyway, .apply() is better for what i am
trying to do here.

TODO convert other np.where uses to .apply

See #4 in https://datatofish.com/if-condition-in-pandas-dataframe/
This commit is contained in:
mlncn 2021-05-02 22:38:54 -04:00
parent d8a5e27a97
commit 206530b235
1 changed files with 10 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
pomodoro_to_harvest.py
View file

@ -1,5 +1,6 @@
import pandas as pd
import numpy as np
import re
import settings
timelog = pd.read_csv(settings.pomodoro_logfile())
@ -47,8 +48,15 @@ timelog['project'] = timelog['project'].str.strip()
# task), then multiply the time by it and remove it from the description.
# Ensure we're splitting on the same asterisk we found: Use the end of string
# signifier in the regular expression ($), and split from the right.
timelog['tmp_multiplier'] = (np.where(timelog['description'].str.contains('\*\s*\d$'), timelog['description'].str.rsplit('*', 1).str[1].str.strip(), 1))
timelog['description'] = (np.where(timelog['description'].str.contains('\*\s*\d$'), timelog['description'].str.rsplit('*', 1).str[0], timelog['description']))
p = re.compile(r'\*\s*\d$')
# On some systems, using np.where worked but others failed. Why it worked is
# unknown but why it failed is because numpy where evaluates all parts, even
# the parts that will never get used because the where clause does not apply!
# This caused the chained strings to fail because— no string.
# timelog['tmp_multiplier'] = (np.where(timelog['description'].str.contains('\*\s*\d$'), timelog['description'].str.rsplit('*', 1).str[1].str.strip(), 1))
# timelog['description'] = (np.where(timelog['description'].str.contains('\*\s*\d$'), timelog['description'].str.rsplit('*', 1).str[0], timelog['description']))
timelog['tmp_multiplier'] = timelog['description'].apply(lambda x: x.rsplit('*', 1)[1].strip() if p.search(x) else 1)
timelog['description'] = timelog['description'].apply(lambda x: x.rsplit('*', 1)[0] if p.search(x) else x)
timelog["time"] = timelog["time"] * timelog['tmp_multiplier'].astype(int)
timelog.drop(columns=['tmp_multiplier'], inplace=True)