is a tool for producing high quality forecasts for time series data that has multiple seasonality with linear or non-linear growth.
| generate_holidays_file.py (prophet-0.7) | : | generate_holidays_file.py (prophet-1.0) |
|---|
| | |
| skipping to change at line 19 | | skipping to change at line 19 |
|---|
| from __future__ import unicode_literals | | from __future__ import unicode_literals |
| | |
| import inspect | | import inspect |
| import unicodedata | | import unicodedata |
| import warnings | | import warnings |
| | |
| import pandas as pd | | import pandas as pd |
| import numpy as np | | import numpy as np |
| | |
| import holidays as hdays_part1 | | import holidays as hdays_part1 |
|
| import fbprophet.hdays as hdays_part2 | | import prophet.hdays as hdays_part2 |
| from fbprophet.make_holidays import make_holidays_df | | from prophet.make_holidays import make_holidays_df |
| | |
| def utf8_to_ascii(text): | | def utf8_to_ascii(text): |
| """Holidays often have utf-8 characters. These are not allowed in R | | """Holidays often have utf-8 characters. These are not allowed in R |
| package data (they generate a NOTE). | | package data (they generate a NOTE). |
| TODO: revisit whether we want to do this lossy conversion. | | TODO: revisit whether we want to do this lossy conversion. |
| """ | | """ |
| ascii_text = ( | | ascii_text = ( |
| unicodedata.normalize('NFD', text) | | unicodedata.normalize('NFD', text) |
| .encode('ascii', 'ignore') | | .encode('ascii', 'ignore') |
| .decode('ascii') | | .decode('ascii') |
| | |
| skipping to change at line 47 | | skipping to change at line 47 |
|---|
| return ascii_text | | return ascii_text |
| | |
| def generate_holidays_file(): | | def generate_holidays_file(): |
| """Generate csv file of all possible holiday names, ds, | | """Generate csv file of all possible holiday names, ds, |
| and countries, year combination | | and countries, year combination |
| """ | | """ |
| year_list = np.arange(1995, 2045, 1).tolist() | | year_list = np.arange(1995, 2045, 1).tolist() |
| all_holidays = [] | | all_holidays = [] |
| # class names in holiday packages which are not countries | | # class names in holiday packages which are not countries |
| # Also cut out countries with utf-8 holidays that don't parse to ascii | | # Also cut out countries with utf-8 holidays that don't parse to ascii |
|
| class_to_exclude = set(['rd', 'BY', 'BG', 'JP', 'RS', 'UA', 'KR']) | | class_to_exclude = {'rd', 'BY', 'BG', 'JP', 'RS', 'UA', 'KR'} |
| | |
| class_list2 = inspect.getmembers(hdays_part2, inspect.isclass) | | class_list2 = inspect.getmembers(hdays_part2, inspect.isclass) |
|
| country_set = set([name for name in list(zip(*class_list2))[0] if len(name)
== 2]) | | country_set = {name for name in list(zip(*class_list2))[0] if len(name) == 2
} |
| class_list1 = inspect.getmembers(hdays_part1, inspect.isclass) | | class_list1 = inspect.getmembers(hdays_part1, inspect.isclass) |
|
| country_set1 = set([name for name in list(zip(*class_list1))[0] if len(name)
== 2]) | | country_set1 = {name for name in list(zip(*class_list1))[0] if len(name) ==
2} |
| country_set.update(country_set1) | | country_set.update(country_set1) |
| country_set -= class_to_exclude | | country_set -= class_to_exclude |
| | |
| for country in country_set: | | for country in country_set: |
| df = make_holidays_df(year_list=year_list, country=country) | | df = make_holidays_df(year_list=year_list, country=country) |
| df['country'] = country | | df['country'] = country |
| all_holidays.append(df) | | all_holidays.append(df) |
| | |
| generated_holidays = pd.concat(all_holidays, axis=0, ignore_index=True) | | generated_holidays = pd.concat(all_holidays, axis=0, ignore_index=True) |
| generated_holidays['year'] = generated_holidays.ds.apply(lambda x: x.year) | | generated_holidays['year'] = generated_holidays.ds.apply(lambda x: x.year) |
| | | |
| End of changes. 4 change blocks. |
|---|
| 5 lines changed or deleted | | 5 lines changed or added |
|---|
"Fossies" - the Fresh Open Source Software Archive 