Related
I am trying to upsample a pandas datetime-indexed dataframe, so that resulting data is equally divided over the new entries.
For instance, let's say I have a dataframe which stores a cost each month, and I want to get a dataframe which summarizes the equivalent costs per day for each month:
df = (pd.DataFrame([[pd.to_datetime('2023-01-01'), 31],
[pd.to_datetime('2023-02-01'), 14]],
columns=['time', 'cost']
)
.set_index("time")
)
Daily costs are 1$ (or whatever currency you like) in January, and 0.5$ in February. My goal in picture:
After a lot of struggle, I managed to obtain the next code snippet which seems to do what I want:
# add a value to perform a correct resampling
df.loc[df.index.max() + relativedelta(months=1)] = 0
# forward-fill over the right scale
# then divide each entry per the number of rows in the month
df = (df
.resample('1d')
.ffill()
.iloc[:-1]
.groupby(lambda x: datetime(x.year, x.month, 1))
.transform(lambda x: (x / x.count()))
)
However, this is not entirely ok:
using transform forces me to have dataframes with a single column ;
I need to hardcode my original frequency several times in different formats (while adding an extra value at the end of the dataframe, and in the groupby), making a function design hard ;
It only works with evenly-spaced datetime index (even if it's ok in my case) ;
it remains complex.
Does anyone have a suggestion to improve that code snippet ?
What if we took df's month indices and expanded them into days range, while dividing df's values by a number those days and assigning to each day, all by list comprehensions (edit: for equally distributed values per day):
import pandas as pd
# initial DataFrame
df = (pd.DataFrame([[pd.to_datetime('2023-01-01'), 31],
[pd.to_datetime('2023-02-01'), 14]],
columns=['time', 'cost']
).set_index("time"))
# reformat to months
df.index = df.index.strftime('%m-%Y')
df1 = pd.concat( # concatenate the resulted DataFrames into one
[pd.DataFrame( # make a DataFrame from a row in df
[v / pd.Period(i).days_in_month # each month's value divided by n of days in a month
for d in range(pd.Period(i).days_in_month)], # repeated for as many times as there are days
index=pd.date_range(start=i, periods=pd.Period(i).days_in_month, freq='D')) # days range
for i, v in df.iterrows()]) # for each df's index and value
df1
Output:
cost
2023-01-01 1.0
2023-01-02 1.0
2023-01-03 1.0
2023-01-04 1.0
2023-01-05 1.0
2023-01-06 1.0
2023-01-07 1.0
2023-01-08 1.0
2023-01-09 1.0
2023-01-10 1.0
2023-01-11 1.0
... ...
2023-02-13 0.5
2023-02-14 0.5
2023-02-15 0.5
2023-02-16 0.5
2023-02-17 0.5
2023-02-18 0.5
2023-02-19 0.5
2023-02-20 0.5
2023-02-21 0.5
2023-02-22 0.5
2023-02-23 0.5
2023-02-24 0.5
2023-02-25 0.5
2023-02-26 0.5
2023-02-27 0.5
2023-02-28 0.5
What could be done to avoid uniform distribution of daily costs and for the cases with multiple columns? Here's an extended df:
# additional columns and a row
df = (pd.DataFrame([[pd.to_datetime('2023-01-01'), 31, 62, 23],
[pd.to_datetime('2023-02-01'), 14, 28, 51],
[pd.to_datetime('2023-03-01'), 16, 33, 21]],
columns=['time', 'cost1', 'cost2', 'cost3']
).set_index("time"))
# reformat to months
df.index = df.index.strftime('%m-%Y')
df
Output:
cost1 cost2 cost3
time
01-2023 31 62 23
02-2023 14 28 51
03-2023 16 33 21
Here's what I came up for the cases where monthly costs may be upsampled by randomized daily costs, inspired by this question. This solution is scalable to the number of columns and rows:
df1 = pd.concat( # concatenate the resulted DataFrames into one
[pd.DataFrame( # make a DataFrame from a row in df
# here we make a Series with random Dirichlet distributed numbers
# with length of a month and a column's value as the sum
[pd.Series((np.random.dirichlet(np.ones(pd.Period(i).days_in_month), size=1)*v
).flatten()) # the product is an ndarray that needs flattening
for v in row], # for every column value in a row
# index renamed as columns because of the created DataFrame's shape
index=df.columns
# transpose and set the proper index
).T.set_index(
pd.date_range(start=i,
periods=pd.Period(i).days_in_month,
freq='D'))
for i, row in df.iterrows()]) # iterate over every row
Output:
cost1 cost2 cost3
2023-01-01 1.703177 1.444117 0.160151
2023-01-02 0.920706 3.664460 0.823405
2023-01-03 1.210426 1.194963 0.294093
2023-01-04 0.214737 1.286273 0.923881
2023-01-05 1.264553 0.380062 0.062829
... ... ... ...
2023-03-27 0.124092 0.615885 0.251369
2023-03-28 0.520578 1.505830 1.632373
2023-03-29 0.245154 3.094078 0.308173
2023-03-30 0.530927 0.406665 1.149860
2023-03-31 0.276992 1.115308 0.432090
90 rows × 3 columns
To assert the monthly sum:
df1.groupby(pd.Grouper(freq='M')).agg('sum')
Output:
cost1 cost2 cost3
2023-01-31 31.0 62.0 23.0
2023-02-28 14.0 28.0 51.0
2023-03-31 16.0 33.0 21.0
I have a dataframe which contains sales information of products, what i need to do is to create a function which based on the product id, product type and date, calculates the average sales for a time period which is less than the given date in the function.
This is how I have implemented it, but this approach takes a lot of time and I was wondering if there was a faster way to do this.
Dataframe:
product_type = ['A','B']
df = pd.DataFrame({'prod_id':np.repeat(np.arange(start=2,stop=5,step=1),235),'prod_type': np.random.choice(np.array(product_type), 705),'sales_time': pd.date_range(start ='1-1-2018',
end ='3-30-2018', freq ='3H'),'sale_amt':np.random.randint(4,100,size = 705)})
Current code:
def cal_avg(product,ptype,pdate):
temp_df = df[(df['prod_id']==product) & (df['prod_type']==ptype) & (df['sales_time']<= pdate)]
return temp_df['sale_amt'].mean()
Calling the function:
cal_avg(2,'A','2018-02-12 15:00:00')
53.983
If you are running the calc_avg function "rarely" then I suggest ignoring my answer. Otherwise, it might be beneficial to you to simply calculate the expanding window average for each product/product type. It might be slow depending on your dataset size (in which case maybe just run it on specific product types?), but you'll only need to run it once. First sort by the column you want to perform the 'expanding' on (expanding is missing the 'on' parameter) to ensure the proper row order. Then 'groupby' and transform each group (to keep the indices of the original dataframe) with your expanding window aggregation of choice (in this case 'mean').
df = df.sort_values('sales_time')
df['exp_mean_sales'] = df.groupby(['prod_id', 'prod_type'])['sale_amt'].transform(lambda gr: gr.expanding().mean())
With the result being:
df.head()
prod_id prod_type sales_time sale_amt exp_mean_sales
0 2 B 2018-01-01 00:00:00 8 8.000000
1 2 B 2018-01-01 03:00:00 72 40.000000
2 2 B 2018-01-01 06:00:00 33 37.666667
3 2 A 2018-01-01 09:00:00 81 81.000000
4 2 B 2018-01-01 12:00:00 83 49.000000
Check Below code, with %%timeit comparison (Google Colab)
import pandas as pd
product_type = ['A','B']
df = pd.DataFrame({'prod_id':np.repeat(np.arange(start=2,stop=5,step=1),235),'prod_type': np.random.choice(np.array(product_type), 705),'sales_time': pd.date_range(start ='1-1-2018',
end ='3-30-2018', freq ='3H'),'sale_amt':np.random.randint(4,100,size = 705)})
## OP's function
def cal_avg(product,ptype,pdate):
temp_df = df[(df['prod_id']==product) & (df['prod_type']==ptype) & (df['sales_time']<= pdate)]
return temp_df['sale_amt'].mean()
## Numpy data prep
prod_id_array = np.array(df.values[:,:1])
prod_type_array = np.array(df.values[:,1:2])
sales_time_array = np.array(df.values[:,2:3], dtype=np.datetime64)
values = np.array(df.values[:,3:])
OP's function -
%%timeit
cal_avg(2,'A','2018-02-12 15:00:00')
Output:
Numpy version
%%timeit -n 1000
cal_vals = [2,'A','2018-02-12 15:00:00']
mask = np.logical_and(prod_id_array == cal_vals[0], prod_type_array == cal_vals[1], sales_time_array <= np.datetime64(cal_vals[2]) )
np.mean(values[mask])
Output:
My data can have multiple events on a given date or NO events on a date. I take these events, get a count by date and plot them. However, when I plot them, my two series don't always match.
idx = pd.date_range(df['simpleDate'].min(), df['simpleDate'].max())
s = df.groupby(['simpleDate']).size()
In the above code idx becomes a range of say 30 dates. 09-01-2013 to 09-30-2013
However S may only have 25 or 26 days because no events happened for a given date. I then get an AssertionError as the sizes dont match when I try to plot:
fig, ax = plt.subplots()
ax.bar(idx.to_pydatetime(), s, color='green')
What's the proper way to tackle this? Do I want to remove dates with no values from IDX or (which I'd rather do) is add to the series the missing date with a count of 0. I'd rather have a full graph of 30 days with 0 values. If this approach is right, any suggestions on how to get started? Do I need some sort of dynamic reindex function?
Here's a snippet of S ( df.groupby(['simpleDate']).size() ), notice no entries for 04 and 05.
09-02-2013 2
09-03-2013 10
09-06-2013 5
09-07-2013 1
You could use Series.reindex:
import pandas as pd
idx = pd.date_range('09-01-2013', '09-30-2013')
s = pd.Series({'09-02-2013': 2,
'09-03-2013': 10,
'09-06-2013': 5,
'09-07-2013': 1})
s.index = pd.DatetimeIndex(s.index)
s = s.reindex(idx, fill_value=0)
print(s)
yields
2013-09-01 0
2013-09-02 2
2013-09-03 10
2013-09-04 0
2013-09-05 0
2013-09-06 5
2013-09-07 1
2013-09-08 0
...
A quicker workaround is to use .asfreq(). This doesn't require creation of a new index to call within .reindex().
# "broken" (staggered) dates
dates = pd.Index([pd.Timestamp('2012-05-01'),
pd.Timestamp('2012-05-04'),
pd.Timestamp('2012-05-06')])
s = pd.Series([1, 2, 3], dates)
print(s.asfreq('D'))
2012-05-01 1.0
2012-05-02 NaN
2012-05-03 NaN
2012-05-04 2.0
2012-05-05 NaN
2012-05-06 3.0
Freq: D, dtype: float64
One issue is that reindex will fail if there are duplicate values. Say we're working with timestamped data, which we want to index by date:
df = pd.DataFrame({
'timestamps': pd.to_datetime(
['2016-11-15 1:00','2016-11-16 2:00','2016-11-16 3:00','2016-11-18 4:00']),
'values':['a','b','c','d']})
df.index = pd.DatetimeIndex(df['timestamps']).floor('D')
df
yields
timestamps values
2016-11-15 "2016-11-15 01:00:00" a
2016-11-16 "2016-11-16 02:00:00" b
2016-11-16 "2016-11-16 03:00:00" c
2016-11-18 "2016-11-18 04:00:00" d
Due to the duplicate 2016-11-16 date, an attempt to reindex:
all_days = pd.date_range(df.index.min(), df.index.max(), freq='D')
df.reindex(all_days)
fails with:
...
ValueError: cannot reindex from a duplicate axis
(by this it means the index has duplicates, not that it is itself a dup)
Instead, we can use .loc to look up entries for all dates in range:
df.loc[all_days]
yields
timestamps values
2016-11-15 "2016-11-15 01:00:00" a
2016-11-16 "2016-11-16 02:00:00" b
2016-11-16 "2016-11-16 03:00:00" c
2016-11-17 NaN NaN
2016-11-18 "2016-11-18 04:00:00" d
fillna can be used on the column series to fill blanks if needed.
An alternative approach is resample, which can handle duplicate dates in addition to missing dates. For example:
df.resample('D').mean()
resample is a deferred operation like groupby so you need to follow it with another operation. In this case mean works well, but you can also use many other pandas methods like max, sum, etc.
Here is the original data, but with an extra entry for '2013-09-03':
val
date
2013-09-02 2
2013-09-03 10
2013-09-03 20 <- duplicate date added to OP's data
2013-09-06 5
2013-09-07 1
And here are the results:
val
date
2013-09-02 2.0
2013-09-03 15.0 <- mean of original values for 2013-09-03
2013-09-04 NaN <- NaN b/c date not present in orig
2013-09-05 NaN <- NaN b/c date not present in orig
2013-09-06 5.0
2013-09-07 1.0
I left the missing dates as NaNs to make it clear how this works, but you can add fillna(0) to replace NaNs with zeroes as requested by the OP or alternatively use something like interpolate() to fill with non-zero values based on the neighboring rows.
Here's a nice method to fill in missing dates into a dataframe, with your choice of fill_value, days_back to fill in, and sort order (date_order) by which to sort the dataframe:
def fill_in_missing_dates(df, date_col_name = 'date',date_order = 'asc', fill_value = 0, days_back = 30):
df.set_index(date_col_name,drop=True,inplace=True)
df.index = pd.DatetimeIndex(df.index)
d = datetime.now().date()
d2 = d - timedelta(days = days_back)
idx = pd.date_range(d2, d, freq = "D")
df = df.reindex(idx,fill_value=fill_value)
df[date_col_name] = pd.DatetimeIndex(df.index)
return df
You can always just use DataFrame.merge() utilizing a left join from an 'All Dates' DataFrame to the 'Missing Dates' DataFrame. Example below.
# example DataFrame with missing dates between min(date) and max(date)
missing_df = pd.DataFrame({
'date':pd.to_datetime([
'2022-02-10'
,'2022-02-11'
,'2022-02-14'
,'2022-02-14'
,'2022-02-24'
,'2022-02-16'
])
,'value':[10,20,5,10,15,30]
})
# first create a DataFrame with all dates between specified start<-->end using pd.date_range()
all_dates = pd.DataFrame(pd.date_range(missing_df['date'].min(), missing_df['date'].max()), columns=['date'])
# from the all_dates DataFrame, left join onto the DataFrame with missing dates
new_df = all_dates.merge(right=missing_df, how='left', on='date')
s.asfreq('D').interpolate().asfreq('Q')
My data can have multiple events on a given date or NO events on a date. I take these events, get a count by date and plot them. However, when I plot them, my two series don't always match.
idx = pd.date_range(df['simpleDate'].min(), df['simpleDate'].max())
s = df.groupby(['simpleDate']).size()
In the above code idx becomes a range of say 30 dates. 09-01-2013 to 09-30-2013
However S may only have 25 or 26 days because no events happened for a given date. I then get an AssertionError as the sizes dont match when I try to plot:
fig, ax = plt.subplots()
ax.bar(idx.to_pydatetime(), s, color='green')
What's the proper way to tackle this? Do I want to remove dates with no values from IDX or (which I'd rather do) is add to the series the missing date with a count of 0. I'd rather have a full graph of 30 days with 0 values. If this approach is right, any suggestions on how to get started? Do I need some sort of dynamic reindex function?
Here's a snippet of S ( df.groupby(['simpleDate']).size() ), notice no entries for 04 and 05.
09-02-2013 2
09-03-2013 10
09-06-2013 5
09-07-2013 1
You could use Series.reindex:
import pandas as pd
idx = pd.date_range('09-01-2013', '09-30-2013')
s = pd.Series({'09-02-2013': 2,
'09-03-2013': 10,
'09-06-2013': 5,
'09-07-2013': 1})
s.index = pd.DatetimeIndex(s.index)
s = s.reindex(idx, fill_value=0)
print(s)
yields
2013-09-01 0
2013-09-02 2
2013-09-03 10
2013-09-04 0
2013-09-05 0
2013-09-06 5
2013-09-07 1
2013-09-08 0
...
A quicker workaround is to use .asfreq(). This doesn't require creation of a new index to call within .reindex().
# "broken" (staggered) dates
dates = pd.Index([pd.Timestamp('2012-05-01'),
pd.Timestamp('2012-05-04'),
pd.Timestamp('2012-05-06')])
s = pd.Series([1, 2, 3], dates)
print(s.asfreq('D'))
2012-05-01 1.0
2012-05-02 NaN
2012-05-03 NaN
2012-05-04 2.0
2012-05-05 NaN
2012-05-06 3.0
Freq: D, dtype: float64
One issue is that reindex will fail if there are duplicate values. Say we're working with timestamped data, which we want to index by date:
df = pd.DataFrame({
'timestamps': pd.to_datetime(
['2016-11-15 1:00','2016-11-16 2:00','2016-11-16 3:00','2016-11-18 4:00']),
'values':['a','b','c','d']})
df.index = pd.DatetimeIndex(df['timestamps']).floor('D')
df
yields
timestamps values
2016-11-15 "2016-11-15 01:00:00" a
2016-11-16 "2016-11-16 02:00:00" b
2016-11-16 "2016-11-16 03:00:00" c
2016-11-18 "2016-11-18 04:00:00" d
Due to the duplicate 2016-11-16 date, an attempt to reindex:
all_days = pd.date_range(df.index.min(), df.index.max(), freq='D')
df.reindex(all_days)
fails with:
...
ValueError: cannot reindex from a duplicate axis
(by this it means the index has duplicates, not that it is itself a dup)
Instead, we can use .loc to look up entries for all dates in range:
df.loc[all_days]
yields
timestamps values
2016-11-15 "2016-11-15 01:00:00" a
2016-11-16 "2016-11-16 02:00:00" b
2016-11-16 "2016-11-16 03:00:00" c
2016-11-17 NaN NaN
2016-11-18 "2016-11-18 04:00:00" d
fillna can be used on the column series to fill blanks if needed.
An alternative approach is resample, which can handle duplicate dates in addition to missing dates. For example:
df.resample('D').mean()
resample is a deferred operation like groupby so you need to follow it with another operation. In this case mean works well, but you can also use many other pandas methods like max, sum, etc.
Here is the original data, but with an extra entry for '2013-09-03':
val
date
2013-09-02 2
2013-09-03 10
2013-09-03 20 <- duplicate date added to OP's data
2013-09-06 5
2013-09-07 1
And here are the results:
val
date
2013-09-02 2.0
2013-09-03 15.0 <- mean of original values for 2013-09-03
2013-09-04 NaN <- NaN b/c date not present in orig
2013-09-05 NaN <- NaN b/c date not present in orig
2013-09-06 5.0
2013-09-07 1.0
I left the missing dates as NaNs to make it clear how this works, but you can add fillna(0) to replace NaNs with zeroes as requested by the OP or alternatively use something like interpolate() to fill with non-zero values based on the neighboring rows.
Here's a nice method to fill in missing dates into a dataframe, with your choice of fill_value, days_back to fill in, and sort order (date_order) by which to sort the dataframe:
def fill_in_missing_dates(df, date_col_name = 'date',date_order = 'asc', fill_value = 0, days_back = 30):
df.set_index(date_col_name,drop=True,inplace=True)
df.index = pd.DatetimeIndex(df.index)
d = datetime.now().date()
d2 = d - timedelta(days = days_back)
idx = pd.date_range(d2, d, freq = "D")
df = df.reindex(idx,fill_value=fill_value)
df[date_col_name] = pd.DatetimeIndex(df.index)
return df
You can always just use DataFrame.merge() utilizing a left join from an 'All Dates' DataFrame to the 'Missing Dates' DataFrame. Example below.
# example DataFrame with missing dates between min(date) and max(date)
missing_df = pd.DataFrame({
'date':pd.to_datetime([
'2022-02-10'
,'2022-02-11'
,'2022-02-14'
,'2022-02-14'
,'2022-02-24'
,'2022-02-16'
])
,'value':[10,20,5,10,15,30]
})
# first create a DataFrame with all dates between specified start<-->end using pd.date_range()
all_dates = pd.DataFrame(pd.date_range(missing_df['date'].min(), missing_df['date'].max()), columns=['date'])
# from the all_dates DataFrame, left join onto the DataFrame with missing dates
new_df = all_dates.merge(right=missing_df, how='left', on='date')
s.asfreq('D').interpolate().asfreq('Q')
I have the data set of customers with their policies, I am trying to find the number of months the customer is with us. (tenure)
df
cust_no poly_no start_date end_date
1 1 2016-06-01 2016-08-31
1 2 2017-05-01 2018-05-31
1 3 2016-11-01 2018-05-31
output should look like,
cust_no no_of_months
1 22
So basically, it should get rid of the months where there is no policy and count the overlapping period once not twice. I have to do this for every customers, so group by cust_no, how can i do this?
Thanks.
One way to do this is to create date ranges for each records, then use stack to get all the months. Next, take the unique values only to count a month only once:
s = df.apply(lambda x: pd.Series(pd.date_range(x.start_date, x.end_date, freq='M').values), axis=1)
ss = s.stack().unique()
ss.shape[0]
Output:
22
For multiple customers you can use groupby. Continuing with #ScottBoston's answer:
df_range = df.apply(lambda r: pd.Series(
pd.date_range(start=r.start_date, end=r.end_date, freq='M')
.values), axis=1)
df_range.groupby('cust_no').apply(lambda x: x.stack().unique().shape[0])