Pandas
Pandas
import pandas as pd
create
# empty
df = pd.DataFrame()
# from 1 dim list
df = pd.DataFrame(data={'Deaths_Covid_2020': l})
# from N dim list, and set name of columns
df = pd.DataFrame(data=l_results, columns=('num_songs_played', 'pct_all_played', 'pct_80pct_played'))
# from multiple lists
data = zip(l_dates, l_deaths2016, l_deaths2017,
l_deaths2018, l_deaths2019, l_deaths2020)
df = pd.DataFrame(data, columns=['Day', '2016', '2017',
'2018', '2019', '2020'])
from csv
see [1]
df = pd.read_csv('data/de-states/de-state-DE-total.tsv', sep="\t")
# only selected columns and parse date
df = pd.read_csv('data/de-states/de-state-DE-total.tsv',
sep="\t",
decimal=",",
usecols=["Date", "Deaths_New",], # only load these columns
parse_dates=["Date",], # convert to date object if format is yyyy-mm-dd pr dd.mm.yyyy
index_col="Date", # choose this column as index
# na_values=["<4",] # values to treat as NA
)
df.rename(columns={"Deaths_New": "Deaths_Covid"}, inplace = True)
# convert data upon loading
# read only first 10 chars from 2021-04-29T12:15:00+02:00 -> 2021-04-29
pd_date_converter = lambda x: (x[0:10])
df = pd.read_csv(
filename,
sep=",",
converters={"Datum": pd_date_converter},
parse_dates=[
"Datum", # # convert to date object if format is yyyy-mm-dd
],
)
# convert datetime to date after loading
for c in ("Buchungstag", "Valutadatum"):
df2[c] = pd.to_datetime(df2[c], format="%d.%m.%Y").dt.date
from list[str] via read_csv
from io import StringIO
csvStringIO = StringIO("\n".join(lines))
df = pd.read_csv(
csvStringIO,
sep=";",
)
from Excel
see [2]
import openpyxl # pip install openpyxl
df = pd.read_excel(
excelFile, sheet_name="COVID_Todesfälle_KW_AG10", engine="openpyxl"
)
df = pd.read_excel(
"in/PP_APLATZ.xlsx",
skiprows=4,
usecols=[
"ArbPlatz",
"BezArt",
],
)
from dict
see [3]
d = {}
for col in df_rki.columns:
d[col] = df_rki[col].sum()
df = pd.DataFrame.from_dict(d, orient="index", columns=["Covid_Tote"])
# deep copy
df2 = df.copy()
from array/record/list
see [4]
read JSON
see [5]
df = pd.read_json("file.json")
export
to csv
df.to_csv(
"data/de-mortality-timeseries.tsv",
sep="\t",
lineterminator="\n",
)
# selected columns only
df[ ["Deaths", "Deaths_roll_av"] ].to_csv(
"data/de-mortality-timeseries.tsv",
sep="\t",
lineterminator="\n",
)
to Excel
see [6]
df.to_excel("data.xlsx", index=False)
Export 2 df as sheets in 1 file
with pd.ExcelWriter('output.xlsx') as writer:
df1.to_excel(writer, sheet_name='Sheet_name_1')
df2.to_excel(writer, sheet_name='Sheet_name_2')
writer.close()
to HTML
see [7]
to JSON
json_data = df.values.tolist()
with Path(pathStatsExport / "ts_types_month.json").open(
"w", encoding="UTF-8"
) as fh:
json.dump(
json_data,
fp=fh,
ensure_ascii=False,
sort_keys=False,
indent=2,
)
Add row
# add dummy row for missing 1.1.2020 df.loc[-1] = "2020-01-01", 0 df.index = df.index + 1 # shifting index df = df.sort_index() # sorting by index
Loop over rows
for index, row in df.iterrows():
selecting
# select columns df = df [ list_of_column_names ] df = df [ [ "date", "gemeindeschluessel", "betten_frei", "betten_belegt" ] ]
substring
df["Was"] = df["Was"].str.slice(0, 4)
sum
# calc sum of row sum_cases = df["Cases"].sum() # sum per index df_sums = df.sum(axis="index") # sum per column df_sums = df.sum(axis="columns")
filter
# filter on column value via str
df = df[df["Date"] == "2021-11-13"]
df = df[~(df["Date"] == "2021-11-13")] # ~ -> invert
# filter on date column via year
df = df[df["Date"].dt.year == 2020]
# multiple
df = df[~df["Date"].isin(("2020-02-29", "2024-02-29", "2028-02-29"))]
# filter on date
df = df[
df["RECEIVE_DATE"].dt.date # convert pandas datetime64[ns] to date
>= (dt.date.today() - dt.timedelta(days=30))
]
# filter on 2 columns
df = df[(df["type"] == "SleepAnalysis") & (df["sourceName"] == "T-iP8")]
df = df[(df["status"] != "CANCELLED") & (df["status"] != "EXPIRED")]
# filter column value based on list
df = df[~((df["Sg"] == 19) & (df["Usg"].isin([0, 1, 2, 3])))]
df = df[df["gemeindeschluessel"].isin(l_lkids)]
df = df[df["Date"].isin(("2020-02-29", "2024-02-29", "2028-02-29"))]
# is not in
df = df[~df["Date"].isin(("2020-02-29", "2024-02-29", "2028-02-29"))]
# remove data from latest week, as it might not be complete
df = df[df["YearWeek"] < df["YearWeek"].max()]
# drop value=na/NaN rows
df = df[df["value"].notna()]
Duplicates
df = df.drop_duplicates()
convert to date
# parse date columns
date_cols = ["start_date", "start_date_local", "x_date"]
for col in date_cols:
df_all[col] = pd.to_datetime(df_all[col]) # type: ignore
# when in a function one might get the SettingWithCopyWarning, fix via
df = df.copy()
# ensure to have datetime64[ns] type
df["Date"] = pd.to_datetime(df["Date"])
df["Date"] = pd.to_datetime(df["Date"], format="%Y-%m-%d")
# remove miliseconds
df["Date"] = df["Date"].dt.ceil(freq="s")
# split "week/year" to date
# 51/2021 -> date of sunday
def convert2date(s: str) -> dt.date:
year, week = s.split("/")
date = dt.date.fromisocalendar(int(week), int(year), 7)
return date
df["dt"] = df["Kalenderwoche"].apply(lambda x: convert2date(x))
df["Year"] = df["Date"].dt.year
Remove timezone offset info
# V1: proper aproach using tz_convert(None)
df[col] = pd.to_datetime(df[col], format="%Y-%m-%dT%H:%M:%S%z")
df[col] = df[col].dt.tz_convert(None)
# if it throws:
# AttributeError: Can only use .dt accessor with datetimelike values. Did you mean: 'at'?
# see V2
# V2: simple removing the timezone offset
# Remove the timezone information by replacing the "+01:00", "+02:00", "-02:00",
# etc. with an empty string
df[col] = df[col].str.replace(r"[+\-]\d{2}:\d{2}.*$", "", regex=True)
# Parse the datetime column without timezone information
df[col] = pd.to_datetime(df[col], format="%Y-%m-%dT%H:%M:%S")
# note: now without the timezone %z info: format="%Y-%m-%dT%H:%M:%S%z"
convert Pandas Timestamp to Python datetime
def convert_pd_timestamp_to_dt(
pd_timestamp: pd.Timestamp,
) -> dt.datetime:
my_date_time = pd_timestamp.to_pydatetime()
return my_date_time
df["Date"] = df["Timestamp"].apply(convert_pd_timestamp_to_dt)
convert date to timestamp
df["timestamp"] = df["date"].values.astype(np.int64) // 10**9
TimeZone conversion
# convert UTC to local time
df["Date"] = (
df["Date"]
.dt.tz_localize("utc")
.dt.tz_convert(tz="Europe/Berlin")
.dt.tz_localize(None) # drop timezone info, since Excel can not handle it
)
or
# convert UTC to local time
from zoneinfo import ZoneInfo
TZ_ES = ZoneInfo("Europe/Madrid")
df["Date"] = df["Date"].tz_convert(tz=TZ_ES)
overwrite data
# overwrite large values
df["time to fall asleep"].where(df["time to fall asleep"] > 100, 100, inplace=True)
# overwrite column data of last 3 weeks by None
df["DateAsDate"] = pd.to_datetime(df["Date"], format="%Y-%m-%d")
date_3w = dt.date.today() - dt.timedelta(weeks=3)
df.loc[df["DateAsDate"].dt.date >= date_3w, "MyColumn"] = None
# overwrite column value based on date index
df.loc[df.index < pd.Timestamp("2020-03-01"), "InzidenzChange"] = None
# rolling takes NAN values into account, so I need to overwrite them as well
df3["Deaths_Covid_roll"] = np.where(
df3["Deaths_Covid"].isnull(), np.nan, df3["Deaths_Covid_roll"]
# replace na by 0
df = df.fillna(0)
# remove word "Probable" from scorpio_call for better clustering
df["scorpio_call"] = df["scorpio_call"].replace(
to_replace=r"^Probable ", value="", regex=True
)
# Add/Overwrite since cell, indexed by datetime
# append today using yesterdays value
str_today = dt.datetime.today().strftime("%Y-%m-%d")
ts_today = pd.Timestamp(str_today)
if df.index[-1] != ts_today:
df.loc[ts_today] = df.loc[df.index[-1], "Count"]
change value based on condition
df.loc[(df["Kat2"] == "") & (df["IBAN"] == "DE02100100100152517108"), "Kat2"] = "Bahn"
df.loc[(df["Kat1"] == "") & (df["Kat2"] == "Bahn"), "Kat1"] = "Mobilität"
# assigning multiple values
df.loc[
(df["Kat1"] == "") & (df["IBAN"] == "DE02100100100152517108"),
["Kat1", "Kat2", "Kat3"],
] = ["Mobilität], "Reisen", "Bahn"]
# str endswith
df.loc[
(df["RespTime"] == 0.0)
& (df["Text"].str.endswith("(read timeout=30)")),
"RespTime",
] = 30.0
split str date
df [ [ "week", "year" ] ] = df["Kalenderwoche"].str.split("/", expand=True)
Sorting / order by
# sort by column df = df.sort_values(by=['betten_belegt'], ascending=False) df = df.sort_values(by=["Buchungstag", "Text"], ignore_index=True)
Group By
# group and count
df1 = df.groupby(['type'])['type'].count()
df1 = df.groupby(['type'])['type'].count().sort_values(ascending=False)
df_lineages = (
df_all_data.groupby(["lineage", "RECEIVE_DATE"]).size().reset_index(name="count")
)
# group and sum all other columns
df = df.groupby(["YearWeek"]).sum()
df = df.groupby(['Impfdatum'])['Anzahl'].sum().reset_index()
df = df.groupby(['date', 'bundesland']).agg(
{'faelle_covid_aktuell_invasiv_beatmet': 'sum',
'betten_ges': 'sum'}
)
df_top_ten = (
df.groupby("myColumn")
.sum() # sums over remaining column "count"
.sort_values(by="count", ascending=False)
.head(10)
)
Datetime Index Grouping
df = df.groupby(pd.Grouper(freq="5min", offset="00h00min")).max()
Group by Type and Date and define different aggregations per column and add missing values
# Group by type and date
df_month = df.groupby(["type", pd.Grouper(key="date", freq="MS")]).agg(
{"id": "count", "minutes": "sum"}
)
# generate index of the desired month-freq:
idx = pd.date_range(
start=df["date"].min().replace(day=1),
end=df["date"].max().replace(day=1),
freq="MS", # MS = Month Start
)
# add missing months
df_month = df_month.reindex(
pd.MultiIndex.from_product(
[df_month.index.get_level_values("type"), idx],
names=["type", "date"],
)
)
# fill missing by 0 and convert count back to type int
df_month = df_month.fillna(0).astype({"count": int})
Group Datetime by Time of Day only
df["Time"] = df["Date"].dt.round("5min").dt.time
df_grouped = df [ ["RespTime", "Time" ] ].groupby("Time").mean()
Group and calculate min,avg,max
df_grouped = (
df[ ["Time", "RespTime"] ]
.groupby("Time")
.agg(
max=pd.NamedAgg(column="RespTime", aggfunc="max"),
avg=pd.NamedAgg(column="RespTime", aggfunc="mean"),
min=pd.NamedAgg(column="RespTime", aggfunc="min"),
)
)
Column Handling
df = df.rename(
columns = {"invasiv_beatmet": "beatmet",},
errors="raise",
)
drop columns
df = df.drop(columns=["Sterbejahr", "Sterbewoche"])
string to float: 1.234,56 -> 1234.56
df["Euro"] = (
df["Euro"]
.str.replace(".", "")
.str.replace(",", ".")
.astype(float)
)
convert int to str adding leading zeros
df["Sterbewoche"].astype(str).str.zfill(2)
sum up over all columns
sum_cases = 0
for col in df.columns:
sum_cases += df[col].sum()
Column Header Handling
# rename column headers by extracting some int values from a string
l2 = []
for col in df.columns:
year = int(col[0:4])
week = int(col[5:7])
l2.append(year * 100 + week)
df.columns = l2
Change Column Order
df = df.reindex(
[
"Wann",
"Art",
"Wer",
],
axis="columns",
)
# move certain columns to front, leaving the remaining unchanged
first_colmns = ("description", "isin", "side", "quantity", "Price", "amount")
i = 0
for col in first_colmns:
df.insert(i, col, df.pop(col))
i += 1
del i
Row Handling
add row from list and index
idx = df.index[-1] + 1 list_of_values = (...) df.loc[idx] = list_of_values
Index Handling
# set index name df.index.name = "YearWeek"
int indexes
select column as index
df.set_index("Altersgruppe", inplace=True)
# reset index to start at 0
df2 = df1[1 * 365 : 2 * 365].reset_index(drop=True)
# filter on index
df = df[df.index >= start_yearweek]
datetime indexes
# select "Date" column as index
df.set_index(["Date"], inplace=True)
# convert index to datetime
df.index = pd.to_datetime(df.index)
# convert utc to local time
df = df.tz_convert("Europe/Berlin")
# remove timezone offset (for Excel can not handle this)
df.index = df.index.tz_localize(None)
# move date from index back to column
df.reset_index(inplace=True)
assert df.columns[0] == "Date"
# filter on date
# drop data prior to 2020
df = df.loc['2020-01-01':]
# alternative:
df = df[df.index >= "2021-01-10"]
date_last = pd.to_datetime(df.index[-1]).date()
# reindex and fill missing with 0
date_last = pd.to_datetime(df.index[-1]).date()
idx = pd.date_range('2020-01-01', date_last))
df = df.reindex(idx, fill_value=0)
# add missing dates
df = df.asfreq('D', fill_value=0)
df = df.asfreq(freq="5M")
df.sort_index(inplace=True) # needed?
# drop values of column for last 3 weeks
date_3w = dt.date.today() - dt.timedelta(weeks=3)
df.loc[df.index.date >= date_3w, "Cases"] = None
# or
df.loc[df.index.date < pd.to_datetime("2020-03-01"), "Value"] = None
text indexes
df.set_index("Altersgruppe", inplace=True)
extract cell
# select column for index
betten_ges = df2["betten_ges"].iloc[0]
# extract cell
de_sum = df["Personen"].loc["Summe"]
# drop row based on index
df.drop("Summe", inplace=True)
melt/unpivot wide table to long table format
df2 = pd.melt(df, id_vars="Date", value_vars=df.columns[1:]) #
df2.rename(columns={"variable": "Machine", "value": "Status"}, inplace=True)
# not working, using melt instead
# df3 = pd.wide_to_long(df, stubnames="Date", i=df.columns[1:], j="Status")
merge/concat multiple dataframes
df_sum = pd.DataFrame()
df_sum = pd.concat(
[df_sum, df],
ignore_index=True,
)
# join series of 2 df
df_covid_2020 = pd.DataFrame()
df_covid_2020['Deaths_Covid_2020'] = df1['Deaths_Covid_2020'].append(
df2['Deaths_Covid_2020'], ignore_index=True)
add rows
df = pd.concat([df, df2], axis=1) # note: this would require all index of df in df2 present: # df[machine] = df2[machine]
add single row of datetimeindex
# df.columns is "Date", "Count"
ts_today = pd.Timestamp(dt.date.today())
# when index it numeric
if df["Date"].iloc[-1] != ts_today:
df.loc[len(df.index)] = [ts_today, 0]
# wen using datetime index
if df.index[-1] != ts_today:
df.loc[ts_today] = 0
more
df['2016_roll'] = df['2016'].rolling(window=7, min_periods=1).mean().round(1)
df['2017_roll'] = df['2017'].rolling(window=7, min_periods=1).mean().round(1)
df['2018_roll'] = df['2018'].rolling(window=7, min_periods=1).mean().round(1)
df['2019_roll'] = df['2019'].rolling(window=7, min_periods=1).mean().round(1)
df['2020_roll'] = df['2020'].rolling(window=7, min_periods=1).mean().round(1)
# mean value of 4 columns
df['2016_2019_mean'] = df.iloc[:, [1, 2, 3, 4]
].mean(axis=1) # not column 0 = day
df['2016_2019_mean_roll'] = df['2016_2019_mean'].rolling(
window=7, min_periods=1).mean().round(1)
df['2016_2019_roll_max'] = df.iloc[:, [6, 7, 8, 9]].max(axis=1)
df['2016_2019_roll_min'] = df.iloc[:, [6, 7, 8, 9]].min(axis=1)
# prepend empty values to df
# Jan und Feb values are missing for Covid Deaths series, so I need a couple of empty rows
l = [None] * 59
df1 = pd.DataFrame(data={'Deaths_Covid_2020': l})
# ensure first row is from 28.2
assert (df2.iloc[0]['Date'] ==
'2020-02-28'), "Error of start date, expecting 2020-02-28"
# copy
df2['Date'] = df0['Date']
# drop 2 rows from the beginning
df2.drop([0, 1], inplace=True)
# negative -> 0
df[df < 0] = 0
Helpers
def pandas_set_date_index(df, date_column: str):
""" use date as index """
df[date_column] = pd.to_datetime(df[date_column], format='%Y-%m-%d')
df.set_index([date_column], inplace=True)
return df
def pandas_calc_roll_av(df, column: str, days: int = 7):
""" calc rolling average over column """
df[column + '_roll_av'] = df[column].rolling(
window=days, min_periods=1).mean().round(1)
return df
Transpose
# transpose to have yearweek as index df = df.transpose()
String Modifications
str.replace()
df["Text"] = df["Text"].str.replace(
r"^Menke, *Torben *",
"Torben Menke",
regex=True,
)
capitalization
df["Wer"] = df["Wer"].apply(lambda x: string.capwords(x))
trim spaces
df["Wer"] = df["Wer"].str.strip()
count word per row and check if more than one str.count()
df["cnt_Buchungstext"] = df["Text"].str.count("word")
df_search = df[df["cnt_Buchungstext"] != 1]
if len(df_search) > 0:
print(df_search)
del df_search
df = df.drop(columns=["cnt_Buchungstext"])
Plotting via Matplotlib
see Matplotlib