I have a table like this (with more columns):
date,Sector,Value1,Value2
14/03/22,Medical,86,64
14/03/22,Medical,464,99
14/03/22,Industry,22,35
14/03/22,Services,555,843
15/03/22,Services,111,533
15/03/22,Industry,222,169
15/03/22,Medical,672,937
15/03/22,Medical,5534,825
I have created some features like this:
sectorGroup = df.groupby(["date","Sector"])["Value1","Value2"].mean().reset_index()
df = pd.merge(df,sectorGroup,on=["date","Sector"],how="left",suffixes=["","_bySector"])
dateGroupGroup = df.groupby(["date"])["Value1","Value2"].mean().reset_index()
df = pd.merge(df,dateGroupGroup,on=["date"],how="left",suffixes=["","_byDate"])
Now my new df looks like this:
date,Sector,Value1,Value2,Value1_bySector,Value2_bySector,Value1_byDate,Value2_byDate
14/03/22,Medical,86,64,275.0,81.5,281.75,260.25
14/03/22,Medical,464,99,275.0,81.5,281.75,260.25
14/03/22,Industry,22,35,22.0,35.0,281.75,260.25
14/03/22,Services,555,843,555.0,843.0,281.75,260.25
15/03/22,Services,111,533,111.0,533.0,1634.75,616.0
15/03/22,Industry,222,169,222.0,169.0,1634.75,616.0
15/03/22,Medical,672,937,3103.0,881.0,1634.75,616.0
15/03/22,Medical,5534,825,3103.0,881.0,1634.75,616.0
Now, I want to create lag features for Value1_bySector,Value2_bySector,Value1_byDate,Value2_byDate
For example, a new column named Value1_by_Date_lag1 and Value1_bySector_lag1.
And this new column will look like this:
date,Sector,Value1_by_Date_lag1,Value1_bySector_lag1
15/03/22,Services,281.75,555.0
15/03/22,Industry,281.75,22.0
15/03/22,Medical,281.75,275.0
15/03/22,Medical,281.75,275.0
Basically in Value1_by_Date_lag1, the date "15/03" will contain the value "281.75" which is for the date "14/03" (lag of 1 shift).
Basically in Value1_bySector_lag1, the date "15/03" and Sector "Medical" will contain the value "275.0", which is the value for "14/03" and "Medical" rows.
I hope, the question is clear and gave you all the details.
Create a lagged date variable by shifting the date column, and then merge again with dateGroupGroup and sectorGroup using the lagged date instead of the actual date.
df = pd.read_csv(io.StringIO("""date,Sector,Value1,Value2
14/03/22,Medical,86,64
14/03/22,Medical,464,99
14/03/22,Industry,22,35
14/03/22,Services,555,843
15/03/22,Services,111,533
15/03/22,Industry,222,169
15/03/22,Medical,672,937
15/03/22,Medical,5534,825"""))
# Add a lagged date variable
lagged = df.groupby("date")["date"].first().shift()
df = df.join(lagged, on="date", rsuffix="_lag")
# Create date and sector groups and merge them into df, as you already do
sectorGroup = df.groupby(["date","Sector"])[["Value1","Value2"]].mean().reset_index()
df = pd.merge(df,sectorGroup,on=["date","Sector"],how="left",suffixes=["","_bySector"])
dateGroupGroup = df.groupby("date")[["Value1","Value2"]].mean().reset_index()
df = pd.merge(df, dateGroupGroup, on="date",how="left", suffixes=["","_byDate"])
# Merge again, this time matching the lagged date in df to the actual date in sectorGroup and dateGroupGroup
df = pd.merge(df, sectorGroup, left_on=["date_lag", "Sector"], right_on=["date", "Sector"], how="left", suffixes=["", "_by_sector_lag"])
df = pd.merge(df, dateGroupGroup, left_on="date_lag", right_on="date", how="left", suffixes=["", "_by_date_lag"])
# Drop the extra unnecessary columns that have been created in the merge
df = df.drop(columns=['date_by_date_lag', 'date_by_sector_lag'])
This assumes the data is sorted by date - if not you will have to sort before generating the lagged date. It will work whether or not all the dates are consecutive.
I found 1 inefficient solution (slow and memory intensive).
Lag of "date" group
cols = ["Value1_byDate","Value2_byDate"]
temp = df[["date"]+cols]
temp = temp.drop_duplicates()
for i in range(10):
temp.date = temp.date.shift(-1-i)
df = pd.merge(df,temp,on="date",how="left",suffixes=["","_lag"+str(i+1)])
Lag of "date" and "Sector" group
cols = ["Value1_bySector","Value2_bySector"]
temp = df[["date","Sector"]+cols]
temp = temp.drop_duplicates()
for i in range(10):
temp[["Value1_bySector","Value2_bySector"]] = temp.groupby("Sector")["Value1_bySector","Value2_bySector"].shift(1+1)
df = pd.merge(df,temp,on=["date","Sector"],how="left",suffixes=["","_lag"+str(i+1)])
Is there a more simple solution?
I have to use the same function twice. The first when the parameter is df, the second when the parameter is df3. How to do that? The function:
def add(df, df3):
df["timestamp"] = pd.to_datetime(df["timestamp"])
df = df.groupby(pd.Grouper(key = "timestamp", freq = "h")).agg("mean")
price = df["price"]
amount = df["amount"]
return (price * amount) // amount
The double use :
out = []
# This loop will use the add(df) function for every csv and append in a list
for f in csv_files:
df = pd.read_csv(f, header=0)
# Replace empty values with numpy, not sure if usefull, maybe pandas can handle this
df.replace("", np.nan)
#added aggregate DataFrame with new column to list of DataFrames
out.append(add(df))
out2 = []
df3 = pd.Series(dtype=np.float64)
for f in csv_files:
df2 = pd.read_csv(f, header=0)
df3 = pd.concat([df3, df2], ignore_index=True)
out2 = pd.DataFrame(add(df = df3))
out2
I got the error:
TypeError: add() missing 1 required positional argument: 'df3'
The names of the add function have nothing to do with the variable names df and df3 in the rest of the script.
As #garagnoth has stated, you only need one parameter in add. You can call it df, foo or myvariablename: it is not related to nor df, nor df3.
In your case, you can change the add function to the following:
def add(a_dataframe):
# I set the argument name to "a_dataframe" so you can
# see its name is not linked to outside variables
a_dataframe["timestamp"] = pd.to_datetime(a_dataframe["timestamp"])
a_dataframe = a_dataframe.groupby(pd.Grouper(key = "timestamp", freq = "h")).agg("mean")
price = a_dataframe["price"]
amount = a_dataframe["amount"]
return (price * amount) // amount
You can now call this function with df or df3 as the rest of the script already does.
I'm having some trouble creating a DataFrame with some Series. Is there a way to concat them with a for? Because each time I try I only get the last Series in the DF, when I really want it to concat it the columns and not in place.
suma_queries = list()
for query in queries:
cur.execute(query)
schema = lib.get_schema_sql(cursor = cur)
table = lib.get_table_sql(cur)
df = pd.DataFrame(data = table, columns = schema)
suma_queries.append(df.iloc[:,18].sum())
suma_queries = pd.Series(suma_queries)
concat_df = pd.concat([suma_queries], axis=1)
As you see, for each "suma_queries" Series gotten from the for, I try to concatenate it to a dataframe called concat_df, and so on for the next "suma_queries" Series, but in the end, I only get the last Series, because the for replaces the value.
What I want at the end should be a dataframe like:
Series1 Series2 Series3 … SeriesN
s1_1 s2_1 s3_1 sn_1
s1_2 … … …
s1_3 … … …
… … … …
s1_n s2_n s3_n sn_n
where each column is a series.
Please let me know if there is way to do it,
Thanks!!
you should reasign the appended dataframe back to itself in the for loop:
for query in queries:
cur.execute(query)
schema = lib.get_schema_sql(cursor = cur)
table = lib.get_table_sql(cur)
df = pd.DataFrame(data = table, columns = schema)
suma_queries=suma_queries.append(df.iloc[:,18].sum())
suma_queries = pd.Series(suma_queries)
concat_df = pd.concat([suma_queries], axis=1)
I am having trouble reformatting a dataframe.
My input is a day value rows by symbols columns (each symbol has different dates with it's values):
Input
code to generate input
data = [("01-01-2010", 15, 10), ("02-01-2010", 16, 11), ("03-01-2010", 16.5, 10.5)]
labels = ["date", "AAPL", "AMZN"]
df_input = pd.DataFrame.from_records(data, columns=labels)
The needed output is (month row with new row for each month):
Needed output
code to generate output
data = [("01-01-2010","29-01-2010", "AAPL", 15, 20), ("01-01-2010","29-01-2010", "AMZN", 10, 15),("02-02-2010","30-02-2010", "AAPL", 20, 32)]
labels = ['bd start month', 'bd end month','stock', 'start_month_value', "end_month_value"]
df = pd.DataFrame.from_records(data, columns=labels)
Meaning (Pseudo code)
1. for each row take only non nan values to create a new "row" (maybe dictionary with the date as the index and the [stock, value] as the value.
2. take only rows that are business start of month or business end of month.
3. write those rows to a new datatframe.
I have read several posts like this and this and several more.
All treat with dataframe of the same "type" and just resampling while I need to change to structure...
My code so far
# creating the new index with business days
df1 =pd.DataFrame(range(10000), index = pd.date_range(df.iloc[0].name, periods=10000, freq='D'))
from pandas.tseries.offsets import CustomBusinessMonthBegin
from pandas.tseries.holiday import USFederalHolidayCalendar
bmth_us = CustomBusinessMonthBegin(calendar=USFederalHolidayCalendar())
df2 = df1.resample(bmth_us).mean()
# creating the new index interseting my old one (daily) with the monthly index
new_index = df.index.intersection(df2.index)
# selecting only the rows I want
df = df.loc[new_index]
# creating a dict that will be my new dataset
new_dict = collections.OrderedDict()
# iterating over the rows and adding to dictionary
for index, row in df.iterrows():
# print index
date = df.loc[index].name
# values are the not none values
values = df.loc[index][~df.loc[index].isnull().values]
new_dict[date]=values
# from dict to list
data=[]
for key, values in new_dict.iteritems():
for i in range(0, len(values)):
date = key
stock_name = str(values.index[i])
stock_value = values.iloc[i]
row = (key, stock_name, stock_value)
data.append(row)
# from the list to df
labels = ['date','stock', 'value']
df = pd.DataFrame.from_records(data, columns=labels)
df.to_excel("migdal_format.xls")
Current output I get
One big problem:
I only get value of the stock on the start of month day.. I need start and end so I can calculate the stock gain on this month..
One smaller problem:
I am sure this is not the cleanest and fastest code :)
Thanks a lot!
So I have found a way.
looping through each column
groupby month
taking the first and last value I have in that month
calculate return
df_migdal = pd.DataFrame()
for col in df_input.columns[0:]:
stock_position = df_input.loc[:,col]
name = stock_position.name
name = re.sub('[^a-zA-Z]+', '', name)
name = name[0:-4]
stock_position=stock_position.groupby([pd.TimeGrouper('M')]).agg(['first', 'last'])
stock_position["name"] = name
stock_position["return"] = ((stock_position["last"] / stock_position["first"]) - 1) * 100
stock_position.dropna(inplace=True)
df_migdal=df_migdal.append(stock_position)
df_migdal=df_migdal.round(decimals=2)
I tried I way cooler way, but did not know how to handle the ,multi index I got... I needed that for each column, to take the two sub columns and create a third one from some lambda function.
df_input.groupby([pd.TimeGrouper('M')]).agg(['first', 'last'])
I have few categorical columns (description) in my DataFrame df_churn which i'd like to convert to numerical values. And of course I'd like to create a lookup table because i will need to convert them back eventually.
The problem is that every column has a different number of categories so appending to df_categories is not easy and I cant think of any simple way of do so.
Here is what I have so far. It stops after first column, because of the different length.
cat_clmn = ['CLI_REGION','CLI_PROVINCE','CLI_ORIGIN','cli_origin2','cli_origin3', 'ONE_PRD_TYPE_1']
df_categories = pd.DataFrame()
def categorizer(_clmn):
for clmn in cat_clmn:
dict_cat = {key: value for value, key in enumerate(df_churn[clmn].unique())}
df_categories[clmn] = dict_cat.values()
df_categories[clmn + '_key'] = dict_cat.keys()
df_churn[clmn + '_CAT'] = df_churn[clmn].map(dict_cat)
categorizer(cat_clmn)
There is a temporary solution, but I am sure it can be done in a better way.
df_CLI_REGION = pd.DataFrame()
df_CLI_PROVINCE = pd.DataFrame()
df_CLI_ORIGIN = pd.DataFrame()
df_cli_origin2 = pd.DataFrame()
df_cli_origin3 = pd.DataFrame()
df_ONE_PRD_TYPE_1 = pd.DataFrame()
cat_clmn = ['CLI_REGION','CLI_PROVINCE','CLI_ORIGIN','cli_origin2','cli_origin3', 'ONE_PRD_TYPE_1']
df_lst = [df_CLI_REGION,df_CLI_PROVINCE,df_CLI_ORIGIN,df_cli_origin2,df_cli_origin3, df_ONE_PRD_TYPE_1]
def categorizer(_clmn):
for clmn, df in zip(cat_clmn,df_lst):
d = {key: value for value, key in enumerate(df_churn[clmn].unique())}
df[clmn] = d.values()
df[clmn + '_key'] = d.keys()
df_churn[clmn + '_CAT'] = df_churn[clmn].map(d)
categorizer(cat_clmn)