Mailing List Archive: .replace() to replace elements in a Pandas DataFrame

.replace() to replace elements in a Pandas DataFrame

Aug 5, 2020, 3:21 PM

Post #1 of 9 (186 views)

Hi,

I have trouble with .replace() method in a Pandas DataFrame.
My code is something like this:
correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
df = df.replace({'Name' : correction_dic1})
Basically, what I am trying to do here is to replace bowels with Spanish
accent with a plain set of English alphabet in all the entries in the
column called 'Name' in the dataframe, df.

A strange thing is that it worked perfectly in the same Jupyter notebook
this morning.
But all of a sudden, it started not doing the replacement any more.

Could anyone tell what could be the problem?

Thanks
Best
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

python at mrabarnett

Aug 5, 2020, 3:56 PM

Post #2 of 9 (186 views)

Permalink

On 2020-08-05 23:21, Michio Suginoo wrote:
> Hi,
>
> I have trouble with .replace() method in a Pandas DataFrame.
> My code is something like this:
> correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
> df = df.replace({'Name' : correction_dic1})
> Basically, what I am trying to do here is to replace bowels with Spanish
> accent with a plain set of English alphabet in all the entries in the
> column called 'Name' in the dataframe, df.
>
> A strange thing is that it worked perfectly in the same Jupyter notebook
> this morning.
> But all of a sudden, it started not doing the replacement any more.
>
> Could anyone tell what could be the problem?
>
Does that replace characters in strings, or does it replace one string
with another, e.g. replace the string 'á' with the string 'a'?
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

msuginoo at reversalpoint

Aug 5, 2020, 4:06 PM

Post #3 of 9 (186 views)

Permalink

Hi MRAB,

Here is an example:
In the original dataframe, I have something like, 'Agronomía'. And I try to
transform it to 'Agronomia'.
In this case, I try to replace í with i: the former with a Spanish accent,
the latter without.

Thanks
Best

On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com> wrote:

> On 2020-08-05 23:21, Michio Suginoo wrote:
> > Hi,
> >
> > I have trouble with .replace() method in a Pandas DataFrame.
> > My code is something like this:
> > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
> > df = df.replace({'Name' : correction_dic1})
> > Basically, what I am trying to do here is to replace bowels with Spanish
> > accent with a plain set of English alphabet in all the entries in the
> > column called 'Name' in the dataframe, df.
> >
> > A strange thing is that it worked perfectly in the same Jupyter notebook
> > this morning.
> > But all of a sudden, it started not doing the replacement any more.
> >
> > Could anyone tell what could be the problem?
> >
> Does that replace characters in strings, or does it replace one string
> with another, e.g. replace the string 'á' with the string 'a'?
> --
> https://mail.python.org/mailman/listinfo/python-list
>
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

python at mrabarnett

Aug 5, 2020, 4:30 PM

Post #4 of 9 (186 views)

Permalink

On 2020-08-06 00:06, Michio Suginoo wrote:
> Hi MRAB,
>
> Here is an example:
> In the original dataframe, I have something like, 'Agronomía'. And I
> try to transform it to 'Agronomia'.
> In this case, I try to replace í with i: the former with a Spanish
> accent, the latter without.
>
That's what you want it to do, but what does it _actually_ do? Does it
_actually_ replace characters in a string, or only match and replace
entire strings?

If you have a dataframe that contains the strings 'Agronomía' and 'í',
does it leave 'Agronomía' as-is but replace 'í' with 'i'?

On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
<mailto:python@mrabarnett.plus.com>> wrote:
>
> On 2020-08-05 23:21, Michio Suginoo wrote:
> > Hi,
> >
> > I have trouble with .replace() method in a Pandas DataFrame.
> > My code is something like this:
> > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
> > df = df.replace({'Name' : correction_dic1})
> > Basically, what I am trying to do here is to replace bowels with
> Spanish
> > accent with a plain set of English alphabet in all the entries
> in the
> > column called 'Name' in the dataframe, df.
> >
> > A strange thing is that it worked perfectly in the same Jupyter
> notebook
> > this morning.
> > But all of a sudden, it started not doing the replacement any more.
> >
> > Could anyone tell what could be the problem?
> >
> Does that replace characters in strings, or does it replace one
> string
> with another, e.g. replace the string 'á' with the string 'a'?
>
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

msuginoo at reversalpoint

Aug 5, 2020, 5:12 PM

Post #5 of 9 (186 views)

Permalink

Hi MRAB,

Sorry, I simply do not understand the intention of your statement.
If .replace() does not do what I want to do: to replace all ' í ' with 'i'
in any form .
Could you advise me other means to do it?

I am still a learner, so I am not knowledgeable. But, I presume, this sort
of needs--to replace parts of string, not an entire string--should be
common.
So, I would presume, there must be some popular way to do it.
If you know, I would appreciate it if you could advise me.

Thanks.
Best

On Wed, Aug 5, 2020 at 8:38 PM MRAB <python@mrabarnett.plus.com> wrote:

> On 2020-08-06 00:06, Michio Suginoo wrote:
> > Hi MRAB,
> >
> > Here is an example:
> > In the original dataframe, I have something like, 'Agronomía'. And I
> > try to transform it to 'Agronomia'.
> > In this case, I try to replace í with i: the former with a Spanish
> > accent, the latter without.
> >
> That's what you want it to do, but what does it _actually_ do? Does it
> _actually_ replace characters in a string, or only match and replace
> entire strings?
>
> If you have a dataframe that contains the strings 'Agronomía' and 'í',
> does it leave 'Agronomía' as-is but replace 'í' with 'i'?
>
> On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
> <mailto:python@mrabarnett.plus.com>> wrote:
> >
> > On 2020-08-05 23:21, Michio Suginoo wrote:
> > > Hi,
> > >
> > > I have trouble with .replace() method in a Pandas DataFrame.
> > > My code is something like this:
> > > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
> > > df = df.replace({'Name' : correction_dic1})
> > > Basically, what I am trying to do here is to replace bowels with
> > Spanish
> > > accent with a plain set of English alphabet in all the entries
> > in the
> > > column called 'Name' in the dataframe, df.
> > >
> > > A strange thing is that it worked perfectly in the same Jupyter
> > notebook
> > > this morning.
> > > But all of a sudden, it started not doing the replacement any more.
> > >
> > > Could anyone tell what could be the problem?
> > >
> > Does that replace characters in strings, or does it replace one
> > string
> > with another, e.g. replace the string 'á' with the string 'a'?
> >
> --
> https://mail.python.org/mailman/listinfo/python-list
>
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

msuginoo at reversalpoint

Aug 5, 2020, 5:16 PM

Post #6 of 9 (186 views)

Permalink

Hi MRAB,
In addition,. I also found out the following does not work either in my
Jupyter Notebook.

correction_dic2={ "La Boca" : "BOCA", "La Paternal": "PATERNAL", "Villa
General Mitre": "VILLA GRAL. MITRE"}
df = df.replace({'Name' : correction_dic2})

This case, unlike the earlier case, is aiming at replacing entire strings.

Thanks,
Best

On Wed, Aug 5, 2020 at 9:12 PM Michio Suginoo <msuginoo@reversalpoint.com>
wrote:

> Hi MRAB,
>
> Sorry, I simply do not understand the intention of your statement.
> If .replace() does not do what I want to do: to replace all ' í ' with
> 'i' in any form .
> Could you advise me other means to do it?
>
> I am still a learner, so I am not knowledgeable. But, I presume, this sort
> of needs--to replace parts of string, not an entire string--should be
> common.
> So, I would presume, there must be some popular way to do it.
> If you know, I would appreciate it if you could advise me.
>
> Thanks.
> Best
>
>
> On Wed, Aug 5, 2020 at 8:38 PM MRAB <python@mrabarnett.plus.com> wrote:
>
>> On 2020-08-06 00:06, Michio Suginoo wrote:
>> > Hi MRAB,
>> >
>> > Here is an example:
>> > In the original dataframe, I have something like, 'Agronomía'. And I
>> > try to transform it to 'Agronomia'.
>> > In this case, I try to replace í with i: the former with a Spanish
>> > accent, the latter without.
>> >
>> That's what you want it to do, but what does it _actually_ do? Does it
>> _actually_ replace characters in a string, or only match and replace
>> entire strings?
>>
>> If you have a dataframe that contains the strings 'Agronomía' and 'í',
>> does it leave 'Agronomía' as-is but replace 'í' with 'i'?
>>
>> On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
>> <mailto:python@mrabarnett.plus.com>> wrote:
>> >
>> > On 2020-08-05 23:21, Michio Suginoo wrote:
>> > > Hi,
>> > >
>> > > I have trouble with .replace() method in a Pandas DataFrame.
>> > > My code is something like this:
>> > > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
>> > > df = df.replace({'Name' : correction_dic1})
>> > > Basically, what I am trying to do here is to replace bowels with
>> > Spanish
>> > > accent with a plain set of English alphabet in all the entries
>> > in the
>> > > column called 'Name' in the dataframe, df.
>> > >
>> > > A strange thing is that it worked perfectly in the same Jupyter
>> > notebook
>> > > this morning.
>> > > But all of a sudden, it started not doing the replacement any
>> more.
>> > >
>> > > Could anyone tell what could be the problem?
>> > >
>> > Does that replace characters in strings, or does it replace one
>> > string
>> > with another, e.g. replace the string 'á' with the string 'a'?
>> >
>> --
>> https://mail.python.org/mailman/listinfo/python-list
>>
>
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

bob at mellowood

Aug 5, 2020, 5:50 PM

Post #7 of 9 (186 views)

Permalink

I don't think that string.replace() has an option for a dictionary of
replacements ... mind you, it might be nice if it did. In the meantime,
have you tried looping through the dictionary and changing each possible
character?

On Wed, Aug 5, 2020 at 5:12 PM Michio Suginoo <msuginoo@reversalpoint.com>
wrote:

> Hi MRAB,
>
> Sorry, I simply do not understand the intention of your statement.
> If .replace() does not do what I want to do: to replace all ' í ' with 'i'
> in any form .
> Could you advise me other means to do it?
>
> I am still a learner, so I am not knowledgeable. But, I presume, this sort
> of needs--to replace parts of string, not an entire string--should be
> common.
> So, I would presume, there must be some popular way to do it.
> If you know, I would appreciate it if you could advise me.
>
> Thanks.
> Best
>
>
> On Wed, Aug 5, 2020 at 8:38 PM MRAB <python@mrabarnett.plus.com> wrote:
>
> > On 2020-08-06 00:06, Michio Suginoo wrote:
> > > Hi MRAB,
> > >
> > > Here is an example:
> > > In the original dataframe, I have something like, 'Agronomía'. And I
> > > try to transform it to 'Agronomia'.
> > > In this case, I try to replace í with i: the former with a Spanish
> > > accent, the latter without.
> > >
> > That's what you want it to do, but what does it _actually_ do? Does it
> > _actually_ replace characters in a string, or only match and replace
> > entire strings?
> >
> > If you have a dataframe that contains the strings 'Agronomía' and 'í',
> > does it leave 'Agronomía' as-is but replace 'í' with 'i'?
> >
> > On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
> > <mailto:python@mrabarnett.plus.com>> wrote:
> > >
> > > On 2020-08-05 23:21, Michio Suginoo wrote:
> > > > Hi,
> > > >
> > > > I have trouble with .replace() method in a Pandas DataFrame.
> > > > My code is something like this:
> > > > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
> > > > df = df.replace({'Name' : correction_dic1})
> > > > Basically, what I am trying to do here is to replace bowels with
> > > Spanish
> > > > accent with a plain set of English alphabet in all the entries
> > > in the
> > > > column called 'Name' in the dataframe, df.
> > > >
> > > > A strange thing is that it worked perfectly in the same Jupyter
> > > notebook
> > > > this morning.
> > > > But all of a sudden, it started not doing the replacement any
> more.
> > > >
> > > > Could anyone tell what could be the problem?
> > > >
> > > Does that replace characters in strings, or does it replace one
> > > string
> > > with another, e.g. replace the string 'á' with the string 'a'?
> > >
> > --
> > https://mail.python.org/mailman/listinfo/python-list
> >
> --
> https://mail.python.org/mailman/listinfo/python-list
>

--

**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob@mellowood.ca
WWW: http://www.mellowood.ca
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

msuginoo at reversalpoint

Aug 5, 2020, 6:15 PM

Post #8 of 9 (186 views)

Permalink

Hi Bob and MRAB,
Both of you, very kind.
Thanks.
I tried with DataFrame as my original code, but this time, I inserted
regex=True and it worked.
Thank you very much!
Best

On Wed, Aug 5, 2020 at 9:50 PM Bob van der Poel <bob@mellowood.ca> wrote:

> I don't think that string.replace() has an option for a dictionary of
> replacements ... mind you, it might be nice if it did. In the meantime,
> have you tried looping through the dictionary and changing each possible
> character?
>
>
> On Wed, Aug 5, 2020 at 5:12 PM Michio Suginoo <msuginoo@reversalpoint.com>
> wrote:
>
>> Hi MRAB,
>>
>> Sorry, I simply do not understand the intention of your statement.
>> If .replace() does not do what I want to do: to replace all ' í ' with
>> 'i'
>> in any form .
>> Could you advise me other means to do it?
>>
>> I am still a learner, so I am not knowledgeable. But, I presume, this sort
>> of needs--to replace parts of string, not an entire string--should be
>> common.
>> So, I would presume, there must be some popular way to do it.
>> If you know, I would appreciate it if you could advise me.
>>
>> Thanks.
>> Best
>>
>>
>> On Wed, Aug 5, 2020 at 8:38 PM MRAB <python@mrabarnett.plus.com> wrote:
>>
>> > On 2020-08-06 00:06, Michio Suginoo wrote:
>> > > Hi MRAB,
>> > >
>> > > Here is an example:
>> > > In the original dataframe, I have something like, 'Agronomía'. And I
>> > > try to transform it to 'Agronomia'.
>> > > In this case, I try to replace í with i: the former with a Spanish
>> > > accent, the latter without.
>> > >
>> > That's what you want it to do, but what does it _actually_ do? Does it
>> > _actually_ replace characters in a string, or only match and replace
>> > entire strings?
>> >
>> > If you have a dataframe that contains the strings 'Agronomía' and 'í',
>> > does it leave 'Agronomía' as-is but replace 'í' with 'i'?
>> >
>> > On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
>> > <mailto:python@mrabarnett.plus.com>> wrote:
>> > >
>> > > On 2020-08-05 23:21, Michio Suginoo wrote:
>> > > > Hi,
>> > > >
>> > > > I have trouble with .replace() method in a Pandas DataFrame.
>> > > > My code is something like this:
>> > > > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
>> > > > df = df.replace({'Name' : correction_dic1})
>> > > > Basically, what I am trying to do here is to replace bowels with
>> > > Spanish
>> > > > accent with a plain set of English alphabet in all the entries
>> > > in the
>> > > > column called 'Name' in the dataframe, df.
>> > > >
>> > > > A strange thing is that it worked perfectly in the same Jupyter
>> > > notebook
>> > > > this morning.
>> > > > But all of a sudden, it started not doing the replacement any
>> more.
>> > > >
>> > > > Could anyone tell what could be the problem?
>> > > >
>> > > Does that replace characters in strings, or does it replace one
>> > > string
>> > > with another, e.g. replace the string 'á' with the string 'a'?
>> > >
>> > --
>> > https://mail.python.org/mailman/listinfo/python-list
>> >
>> --
>> https://mail.python.org/mailman/listinfo/python-list
>>
>
>
> --
>
> **** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
> Bob van der Poel ** Wynndel, British Columbia, CANADA **
> EMAIL: bob@mellowood.ca
> WWW: http://www.mellowood.ca
>
--
https://mail.python.org/mailman/listinfo/python-list

Re: .replace() to replace elements in a Pandas DataFrame [ In reply to ]

msuginoo at reversalpoint

Aug 5, 2020, 6:17 PM

Post #9 of 9 (186 views)

Permalink

That said, what remains mystery is why the original one worked this morning.
Strange!

On Wed, Aug 5, 2020 at 10:15 PM Michio Suginoo <msuginoo@reversalpoint.com>
wrote:

> Hi Bob and MRAB,
> Both of you, very kind.
> Thanks.
> I tried with DataFrame as my original code, but this time, I inserted
> regex=True and it worked.
> Thank you very much!
> Best
>
>
> On Wed, Aug 5, 2020 at 9:50 PM Bob van der Poel <bob@mellowood.ca> wrote:
>
>> I don't think that string.replace() has an option for a dictionary of
>> replacements ... mind you, it might be nice if it did. In the meantime,
>> have you tried looping through the dictionary and changing each possible
>> character?
>>
>>
>> On Wed, Aug 5, 2020 at 5:12 PM Michio Suginoo <msuginoo@reversalpoint.com>
>> wrote:
>>
>>> Hi MRAB,
>>>
>>> Sorry, I simply do not understand the intention of your statement.
>>> If .replace() does not do what I want to do: to replace all ' í ' with
>>> 'i'
>>> in any form .
>>> Could you advise me other means to do it?
>>>
>>> I am still a learner, so I am not knowledgeable. But, I presume, this
>>> sort
>>> of needs--to replace parts of string, not an entire string--should be
>>> common.
>>> So, I would presume, there must be some popular way to do it.
>>> If you know, I would appreciate it if you could advise me.
>>>
>>> Thanks.
>>> Best
>>>
>>>
>>> On Wed, Aug 5, 2020 at 8:38 PM MRAB <python@mrabarnett.plus.com> wrote:
>>>
>>> > On 2020-08-06 00:06, Michio Suginoo wrote:
>>> > > Hi MRAB,
>>> > >
>>> > > Here is an example:
>>> > > In the original dataframe, I have something like, 'Agronomía'. And I
>>> > > try to transform it to 'Agronomia'.
>>> > > In this case, I try to replace í with i: the former with a Spanish
>>> > > accent, the latter without.
>>> > >
>>> > That's what you want it to do, but what does it _actually_ do? Does it
>>> > _actually_ replace characters in a string, or only match and replace
>>> > entire strings?
>>> >
>>> > If you have a dataframe that contains the strings 'Agronomía' and 'í',
>>> > does it leave 'Agronomía' as-is but replace 'í' with 'i'?
>>> >
>>> > On Wed, Aug 5, 2020 at 8:00 PM MRAB <python@mrabarnett.plus.com
>>> > <mailto:python@mrabarnett.plus.com>> wrote:
>>> > >
>>> > > On 2020-08-05 23:21, Michio Suginoo wrote:
>>> > > > Hi,
>>> > > >
>>> > > > I have trouble with .replace() method in a Pandas DataFrame.
>>> > > > My code is something like this:
>>> > > > correction_dic1 = {'á': 'a', 'í': 'i', 'ú': 'u', 'ó': 'o'}
>>> > > > df = df.replace({'Name' : correction_dic1})
>>> > > > Basically, what I am trying to do here is to replace bowels
>>> with
>>> > > Spanish
>>> > > > accent with a plain set of English alphabet in all the entries
>>> > > in the
>>> > > > column called 'Name' in the dataframe, df.
>>> > > >
>>> > > > A strange thing is that it worked perfectly in the same Jupyter
>>> > > notebook
>>> > > > this morning.
>>> > > > But all of a sudden, it started not doing the replacement any
>>> more.
>>> > > >
>>> > > > Could anyone tell what could be the problem?
>>> > > >
>>> > > Does that replace characters in strings, or does it replace one
>>> > > string
>>> > > with another, e.g. replace the string 'á' with the string 'a'?
>>> > >
>>> > --
>>> > https://mail.python.org/mailman/listinfo/python-list
>>> >
>>> --
>>> https://mail.python.org/mailman/listinfo/python-list
>>>
>>
>>
>> --
>>
>> **** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
>> Bob van der Poel ** Wynndel, British Columbia, CANADA **
>> EMAIL: bob@mellowood.ca
>> WWW: http://www.mellowood.ca
>>
>
--
https://mail.python.org/mailman/listinfo/python-list