from boo import download, read_dataframe download(2012) df = read_dataframe(2012) print(df.head())
"""Download and truncate Rosstat corporate dataset."""
from boo import download, build, read_dataframe, files
print("Please be prepared: "
"download and build operations "
"can take long time!")
year = 2012
# Download raw file from Rosstat
try:
download(year)
except FileExistsError:
print("Raw file already downloaded")
# Select fewer columns and assign short column names
# Will save to new file
try:
build(year)
except FileExistsError:
print("Work file already created")
# Read data as dataframe
df = read_dataframe(year)
print(year, "dataset:", df.shape[0], "rows and", df.shape[1], "columns")
print("File locations:", files(year))
# convert from thousand to billion rub and round to 3 digits
cols = numeric_columns(df)
zf = df.copy()
zf.loc[:, cols] = zf.loc[:, cols].divide(divide_by).round(digits)
return zf
# save as CSV and Excel
def locate(filename):
return os.path.join("assets", filename)
if __name__ == "__main__":
must_overwrite = True
boo.download(2018)
source_df = boo.read_dataframe(2018)
print("Finished reading file, querying...")
# Has some profit or loss, but not exactly zero thousand RUB,
# (protects from ghost firms)
ix = source_df.profit_before_tax != 0
# Not a financial firm
ix = ix & (~source_df.ok1.isin([64, 65]))
# Gazprom will be on top of list
df = source_df[ix].sort_values("ta", ascending=False).dropna()
if must_overwrite:
print("Saving files...")