dataframe transformation of pandas dataframe

Question

I want to manipulate the dataframe element into the data type of the element. the example below dataframe.

A      B         C           D
1      2       12-03-2010   1
45     acxz    3-02-2010    3
43     2.9     14-08-2010   4
45.3   3.5     23-09-2019   2.46
nan    34.23   09-09-2020   09-02-2015
45     nan     sam          10-09-2016
24     45.23   02-03-2021   wax
xyz    32      raz          na

Transform this dataframe into below dataframe.

A        B         C         D
int     int     datetime    int
int     string  datetime    int
int     float   datetime    int
float   float   datetime    float
string  float   datetime    datetime
int     string  string      datetime
int     float   datetime    string
string  int     string      string

can anyone suggest me any way to achieve the below dataframe from the above data using pandas.

@jezrael but head only extract at max 5 data without argument and below one is new dataframe from the above dataframe. — Naruto, Feb 02 '21 at 07:14
so, since `head` is the issue in the example code from jezrael, all you need to do is.... — Pac0, Feb 02 '21 at 07:14

score 1 · Answer 1 · answered Feb 02 '21 at 07:41

1

Use DataFrame.applymap by types, convert to strings and last use replace:

df = pd.DataFrame({
    "a": [1, 0.01, 'aaa', pd.to_datetime('2015-01-01')],
    "b": ['yy', pd.to_datetime('2015-01-01'), 1, 0]
   
})

d = {"<class 'pandas._libs.tslibs.timestamps.Timestamp'>":'datetime',
     "<class 'int'>":'int',
     "<class 'float'>":'float',
     "<class 'str'>":'string'}
df = df.applymap(type).astype(str).replace(d)
print (df)
          a         b
0       int    string
1     float  datetime
2    string       int
3  datetime       int

answered Feb 02 '21 at 07:41

jezrael

822,522
95
1,334
1,252

how you are generating d variable dictonary. – Naruto Feb 02 '21 at 07:53
@Rishav - Manually, it is only for replace – jezrael Feb 02 '21 at 07:54
great @Jezrael. – Naruto Feb 02 '21 at 07:56
@jezreal yes, but need to check for Nan values. – Naruto Feb 02 '21 at 08:03
@jezreal getting issue with DateTime, the above script treats it as a string, not as DateTime. – Naruto Feb 02 '21 at 11:18
@Rishav - it means in input data are datetimes like strings. – jezrael Feb 02 '21 at 11:20
@jezreal to input the data in dataframe I am using read_csv(path,engine='python') it will auto-detect datatime columns of input data. Is there any way to handle the auto conversion? – Naruto Feb 02 '21 at 11:27
@Rishav - No, there is no such option, you need [this](https://stackoverflow.com/questions/17465045/can-pandas-automatically-recognize-dates) for set values manually – jezrael Feb 02 '21 at 11:29

dataframe transformation of pandas dataframe

1 Answers1