I have 3 different columns of timestamps in a pandas dataframe, two of which have fraction of seconds recorded while the third does not have fraction of seconds. I would like to calculate an average of these 3 columns.
I have already tried to compute the average using the mean function on the columns and consistently received nan as the result
import pandas as pd
data = [{'time1': '2018-07-22 04:34:10.8966', 'time2': '2017-07-22 04:34:10.8966', 'time3': '2018-07-27 00:10:04'}]
df = pd.DataFrame(data)
df['estimate'] = df[['time1', 'time2', 'time3']].mean(axis=1)
df
Expected : An average of the 3 timestamps
Actual : While there is no error, it also always evaluates to nan which is not what is desired.
As far as I know you can't to it directly on datetime values, you need to convert them, average, and then convert back:
data = [{'time1': '2018-07-22 04:34:10.8966', 'time2': '2017-07-22 04:34:10.8966', 'time3': '2018-07-27 00:10:04'}]
df = pd.DataFrame(data).apply(pd.to_datetime)
df['estimate'] = pd.to_datetime(df[['time1', 'time2', 'time3']].values.astype(pd.np.int64).mean(axis=1))
Result:
time1 time2 time3 estimate
0 2018-07-22 04:34:10.896600 2017-07-22 04:34:10.896600 2018-07-27 00:10:04 2018-03-24 03:06:08.597733376
Related
I have run into a problem in transforming a dataframe. I'm trying to widen a table grouped on a datetime column, but cant seem to make it work. I have tried to transpose it, and pivot it but cant really make it the way i want it.
Example table:
datetime value
2022-04-29T02:00:00.000000000 5
2022-04-29T03:00:00.000000000 6
2022-05-29T02:00:00.000000000 5
2022-05-29T03:00:00.000000000 7
What I want to achieve is:
index date 02:00 03:00
1 2022-04-29 5 6
2 2022-05-29 5 7
The real data has one data point from 00:00 - 20:00 fore each day. So I guess a loop would be the way to go to generate the columns.
Does anyone know a way to solve this, or can nudge me in the right direction?
Thanks in advance!
Assuming from details you have provided, I think you are dealing with timeseries data and you have data from different dates acquired at 02:00:00 and 03:00:00. Please correct me if I am wrong.
First we replicate your DataFrame object.
import datetime as dt
from io import StringIO
import pandas as pd
data_str = """2022-04-29T02:00:00.000000000 5
2022-04-29T03:00:00.000000000 6
2022-05-29T02:00:00.000000000 5
2022-05-29T03:00:00.000000000 7"""
df = pd.read_csv(StringIO(data_str), sep=" ", header=None)
df.columns = ["date", "value"]
now we calculate unique days where you acquired data:
unique_days = df["date"].apply(lambda x: dt.datetime.strptime(x[:-3], "%Y-%m-%dT%H:%M:%S.%f").date()).unique()
Here I trimmed last 3 0s from your date because it would get complicated to parse. We convert the datetime to datetime object and get unique values
Now we create a new empty df in desired form:
new_df = pd.DataFrame(columns=["date", "02:00", "03:00"])
after this we can populate the values:
for day in unique_days:
new_row_data = [day] # this creates a row of 3 elems, which will be inserted into empty df
new_row_data.append(df.loc[df["date"] == f"{day}T02:00:00.000000000", "value"].values[0]) # here we find data for 02:00 for that date
new_row_data.append(df.loc[df["date"] == f"{day}T03:00:00.000000000", "value"].values[0]) # here we find data for 03:00 same day
new_df.loc[len(new_df)] = new_row_data # now we insert row to last pos
this should give you:
date 02:00 03:00
0 2022-04-29 5 6
1 2022-05-29 5 7
I have CSV file where the second column indicates a time point with the format HHMMSS.
ID;TIME
A;110500
B;090000
C;130200
This situation indicates some questions for me.
Does pandas have a data format to represent a time point with hour, minutes and seconds but without the day, month, ...?
How can I convert that fields to such a format?
On Python I would iterate over the fields. But I am sure that Pandas have a more efficient way.
If there is no time of day format without date I could add a day-month-year date to that timepoint.
That is an MWE
import pandas
import io
csv = io.StringIO('ID;TIME\nA;110500\nB;090000\nC;130200')
df = pandas.read_csv(csv, sep=';')
print(df)
Results in
ID TIME
0 A 110500
1 B 90000
2 C 130200
But what I want to see is
ID TIME
0 A 11:05:00
1 B 9:00:00
2 C 13:02:00
Or much better cutting the seconds also
ID TIME
0 A 11:05
1 B 9:00
2 C 13:02
You could use the parameter date_parser in read_csv like and the time accesor
df = pandas.read_csv(csv, sep=';',
parse_dates=[1], # need to know the position of the TIME column
date_parser=lambda x: pandas.to_datetime(x, format='%H%M%S').time)
print(df)
ID TIME
0 A 11:05:00
1 B 09:00:00
2 C 13:02:00
But doing it after reading might be as good
df = (pandas.read_csv(csv, sep=';')
.assign(TIME=lambda x: pandas.to_datetime(x['TIME'], format='%H%M%S').dt.time)
#or lambda x: pandas.to_datetime(x['TIME'], format='%H%M%S').dt.strftime('%#H:%M')
)
Pandas by default represent dates with datetime64 [ns], so I have in my columns this format [2016-02-05 00:00:00] but I just want the date 2016-02-05, so I applied this code for a few columns:
df3a['MA'] = pd.to_datetime(df3a['MA'])
df3a['BA'] = pd.to_datetime(df3a['BA'])
df3a['FF'] = pd.to_datetime(df3a['FF'])
df3a['JJ'] = pd.to_datetime(df3a['JJ'])
.....
but it gives me as result this error: TypeError: type unhashable: 'numpy.ndarray'
my question is: why i got this error and how do i convert datetime to date for multiple columns (around 50)?
i will be grateful for your help
One way to achieve what you'd like is with a DatetimeIndex. I've first created an Example DataFrame with 'date' and 'values' columns and tried from there on to reproduce the error you've got.
import pandas as pd
import numpy as np
# Example DataFrame with a DatetimeIndex (dti)
dti = pd.date_range('2020-12-01','2020-12-17') # dates from first of december up to date
values = np.random.choice(range(1, 101), len(dti)) # random values between 1 and 100
df = pd.DataFrame({'date':dti,'values':values}, index=range(len(dti)))
print(df.head())
>>> date values
0 2020-12-01 85
1 2020-12-02 100
2 2020-12-03 96
3 2020-12-04 40
4 2020-12-05 27
In the example, just the dates are already shown without the time in the 'date' column, I guess since it is a DatetimeIndex.
What I haven't tested but might can work for you is:
# Your dataframe
df3a['MA'] = pd.DatetimeIndex(df3a['MA'])
...
# automated transform for all columns (if all columns are datetimes!)
for label in df3a.columns:
df3a[label] = pd.DatetimeIndex(df3a[label])
Use DataFrame.apply:
cols = ['MA', 'BA', 'FF', 'JJ']
df3a[cols] = df3a[cols].apply(pd.to_datetime)
I am experimenting with Dask, but I encountered a problem while using apply after grouping.
I have a Dask DataFrame with a large number of rows. Let's consider for example the following
N=10000
df = pd.DataFrame({'col_1':np.random.random(N), 'col_2': np.random.random(N) })
ddf = dd.from_pandas(df, npartitions=8)
I want to bin the values of col_1 and I follow the solution from here
bins = np.linspace(0,1,11)
labels = list(range(len(bins)-1))
ddf2 = ddf.map_partitions(test_f, 'col_1',bins,labels)
where
def test_f(df,col,bins,labels):
return df.assign(bin_num = pd.cut(df[col],bins,labels=labels))
and this works as I expect it to.
Now I want to take the median value in each bin (taken from here)
median = ddf2.groupby('bin_num')['col_1'].apply(pd.Series.median).compute()
Having 10 bins, I expect median to have 10 rows, but it actually has 80. The dataframe has 8 partitions so I guess that somehow the apply is working on each one individually.
However, If I want the mean and use mean
median = ddf2.groupby('bin_num')['col_1'].mean().compute()
it works and the output has 10 rows.
The question is then: what am I doing wrong that is preventing apply from operating as mean?
Maybe this warning is the key (Dask doc: SeriesGroupBy.apply) :
Pandas’ groupby-apply can be used to to apply arbitrary functions, including aggregations that result in one row per group. Dask’s groupby-apply will apply func once to each partition-group pair, so when func is a reduction you’ll end up with one row per partition-group pair. To apply a custom aggregation with Dask, use dask.dataframe.groupby.Aggregation.
You are right! I was able to reproduce your problem on Dask 2.11.0. The good news is that there's a solution! It appears that the Dask groupby problem is specifically with the category type (pandas.core.dtypes.dtypes.CategoricalDtype). By casting the category column to another column type (float, int, str), then the groupby will work correctly.
Here's your code that I copied:
import dask.dataframe as dd
import pandas as pd
import numpy as np
def test_f(df, col, bins, labels):
return df.assign(bin_num=pd.cut(df[col], bins, labels=labels))
N = 10000
df = pd.DataFrame({'col_1': np.random.random(N), 'col_2': np.random.random(N)})
ddf = dd.from_pandas(df, npartitions=8)
bins = np.linspace(0,1,11)
labels = list(range(len(bins)-1))
ddf2 = ddf.map_partitions(test_f, 'col_1', bins, labels)
print(ddf2.groupby('bin_num')['col_1'].apply(pd.Series.median).compute())
which prints out the problem you mentioned
bin_num
0 NaN
1 NaN
2 NaN
3 NaN
4 NaN
...
5 0.550844
6 0.651036
7 0.751220
8 NaN
9 NaN
Name: col_1, Length: 80, dtype: float64
Here's my solution:
ddf3 = ddf2.copy()
ddf3["bin_num"] = ddf3["bin_num"].astype("int")
print(ddf3.groupby('bin_num')['col_1'].apply(pd.Series.median).compute())
which printed:
bin_num
9 0.951369
2 0.249150
1 0.149563
0 0.049897
3 0.347906
8 0.847819
4 0.449029
5 0.550608
6 0.652778
7 0.749922
Name: col_1, dtype: float64
#MRocklin or #TomAugspurger
Would you be able to create a fix for this in a new release? I think there is sufficient reproducible code here. Thanks for all your hard work. I love Dask and use it every day ;)
./test.csv looks like:
price datetime
1 100 2019-10-10
2 150 2019-11-10
...
import pandas as pd
import datetime as date
import datetime as time
from datetime import datetime
from datetime import timedelta
csv_df = pd.read_csv('./test.csv')
today = datetime.today()
csv_df['datetime'] = csv_df['expiration_date'].apply(lambda x: pd.to_datetime(x)) #convert `expiration_date` to datetime Series
def days_until_exp(expiration_date, today):
diff = (expiration_date - today)
return [diff]
csv_df['days_until_expiration'] = csv_df['datetime'].apply(lambda x: days_until_exp(csv_df['datetime'], today))
I am trying to iterate over a specific column in my DateFrame labeled csv_df['datetime'] which in each cell has just one value, a date, and do a calcation defined by diff.
Then I want the single value diff to be put into the new Series csv_df['days_until_expiration'].
The problem is, it's calculating values for every row (673 rows) and putting all those values in a list in each row of csv_df['days_until_expiration. I realize it may be due to the brackets around [diff], but without them I get an error.
In Excel, I would just do something like =SUM(datetime - price) and click and drag down the rows to have it populate a new column. However, I want to do this in Pandas as it's part of a bigger application.
csv_df['datetime'] is series, so x of apply is each cell of series. You call apply with lambda and days_until_exp(), but you doesn't passing x to it. Therefore, the result is wrong.
Anyway, Without your sample data, I guess that you want to find sum of csv_df['datetime'] - today(). To do this, you don't need apply. Just do direct vectorized operation on series and sum.
I make 2 columns dataframe for sample:
csv_df:
datetime days_until_expiration
0 2019-09-01 NaN
1 2019-09-02 NaN
2 2019-09-03 NaN
Do the following return series of delta between csv_df['datetime'] and today(). I guess you want this::
td = datetime.datetime.today()
csv_df['days_until_expiration'] = (csv_df['datetime'] - td).dt.days
csv_df:
datetime days_until_expiration
0 2019-09-01 115
1 2019-09-02 116
2 2019-09-03 117
OR:
To find sum of all deltas and assign the same sum value to csv_df['days_until_expiration']
csv_df['days_until_expiration'] = (csv_df['datetime'] - td).dt.days.sum()
csv_df:
datetime days_until_expiration
0 2019-09-01 348
1 2019-09-02 348
2 2019-09-03 348