Suppose I have n number of data frames df_1, df_2, df_3, ... df_n, containing respectively columns named SPEED1 ,SPEED2, SPEED3, ..., SPEEDn, for instance:
import numpy as np
df_1 = pd.DataFrame({'SPEED1':np.random.uniform(0,600,100)})
df_2 = pd.DataFrame({'SPEED2':np.random.uniform(0,600,100)})
and I want to make the same changes to all of the data frames. How do I do so by defining a function on similar lines?
def modify(df,nr):
df_invalid_nr=df_nr[df_nr['SPEED'+str(nr)]>500]
df_valid_nr=~df_invalid_nr
Invalid_cycles_nr=df[df_invalid]
df=df[df_valid]
print(Invalid_cycles_nr)
print(df)
So, when I try to run the above function
modify(df_1,1)
It returns the entire data frame without modification and the invalid cycles as an empty array. I am guessing I need to define the modification on the global dataframe somewhere in the function for this to work.
I am also not sure if I could do this another way, say just looping an iterator through all the data frames. But, I am not sure it will work.
for i in range(1,n+1):
df_invalid_i=df_i[df_i['SPEED'+str(i)]>500]
df_valid_i=~df_invalid_i
Invalid_cycles_i=df[df_invalid]
df=df[df_valid]
print(Invalid_cycles_i)
print(df)
How do I, in general, access df_1 using an iterator? It seems to be a problem.
Any help would be appreciated, thanks!
Solution
Inputs
import pandas as pd
import numpy as np
df_1 = pd.DataFrame({'SPEED1':np.random.uniform(1,600,100))
df_2 = pd.DataFrame({'SPEED2':np.random.uniform(1,600,100))
Code
To my mind a better approach would be to store your dfs into a list and enumerate over it for augmenting informations into your dfs to create a valid column:
for idx, df in enumerate([df_1, df_2]):
col = 'SPEED'+str(idx+1)
df['valid'] = df[col] <= 500
print(df_1)
SPEED valid
0 516.395756 False
1 14.643694 True
2 478.085372 True
3 592.831029 False
4 1.431332 True
You can then filter for valid or invalid with df_1[df_1.valid] or df_1[df_1.valid == False]
It is a solution to fit your problem, see Another solution that may be more clean and Notes below for explanations you need.
Another (better?) solution
If it is possible for you re-think your code. Each DataFrame has one column speed, then name it SPEED:
dfs = dict(df_1=pd.DataFrame({'SPEED':np.random.uniform(0,600,100)}),
df_2=pd.DataFrame({'SPEED':np.random.uniform(0,600,100)}))
It will allow you to do the following one liner:
dfs = dict(map(lambda key_val: (key_val[0],
key_val[1].assign(valid = key_val[1]['SPEED'] <= 500)),
dfs.items()))
print(dfs['df_1'])
SPEED valid
0 516.395756 False
1 14.643694 True
2 478.085372 True
3 592.831029 False
4 1.431332 True
Explanations:
dfs.items() returns a list of key (i.e. names) and values (i.e. DataFrames)
map(foo, bar) apply the function foo (see this answer, and DataFrame assign) to all the elements of bar (i.e. to all the key/value pairs of dfs.items().
dict() cast the map to a dict.
Notes
About modify
Notice that your function modify is not returning anything... I suggest you to have more readings on mutability and immutability in Python. This article is interesting.
You can then test the following for instance:
def modify(df):
df=df[df.SPEED1<0.5]
#The change in df is on the scope of the function only,
#it will not modify your input, return the df...
return df
#... and affect the output to apply changes
df_1 = modify(df_1)
About access df_1 using an iterator
Notice that when you do:
for i in range(1,n+1):
df_i something
df_i in your loop will call the object df_i for each iteration (and not df_1 etc.)
To call an object by its name, use globals()['df_'+str(i)] instead (Assuming that df_1 to df_n+1 are located in globals()) - from this answer.
To my mind it is not a clean approach. I don't know how do you create your DataFrames but if it is possible for your I will suggest you to store them into a dictionary instead affecting manually:
dfs = {}
dfs['df_1'] = ...
or a bit more automatically if df_1 to df_n already exist - according to first part of vestland answer :
dfs = dict((var, eval(var)) for
var in dir() if
isinstance(eval(var), pd.core.frame.DataFrame) and 'df_' in var)
Then it would be easier for your to iterate over your DataFrames:
for i in range(1,n+1):
dfs['df_'+str(i)'] something
You can use the globals() function which allows you to get a variable by his name.
I just add df_i = globals()["df_"+str(i)] at the begining of the for loop :
for i in range(1,n+1):
df_i = globals()["df_"+str(i)]
df_invalid_i=df_i.loc[df_i['SPEED'+str(i)]>500]
df_valid_i=~df_invalid_i
Invalid_cycles_i=df[df_invalid]
df=df[df_valid]
print(Invalid_cycles_i)
print(df)
Your code sample leaves me a little confused, but focusing on
I want to make the same changes to all of the data frames.
and
How do I, in general, access df_1 using an iterator?
you can do exactly that by organizing your dataframes (dfs) in a dictionary (dict).
Here's how:
Assuming you've got a bunch of variables in your namespace...
# Imports
import pandas as pd
import numpy as np
# A few dataframes with random numbers
# df_1
np.random.seed(123)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_1 = pd.DataFrame(np.random.randint(100,150,size=(rows, 2)), columns=['a', 'b'])
df_1 = df_1.set_index(rng)
# df_2
np.random.seed(456)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_2 = pd.DataFrame(np.random.randint(100,150,size=(rows, 2)), columns=['c', 'd'])
df_2 = df_2.set_index(rng)
# df_3
np.random.seed(789)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_3 = pd.DataFrame(np.random.randint(100,150,size=(rows, 2)), columns=['e', 'f'])
df_3 = df_3.set_index(rng)
...you can identify all that are dataframes using:
alldfs = [var for var in dir() if isinstance(eval(var), pd.core.frame.DataFrame)]
If you've got a lot of different dataframes but would only like to focus on those that have a prefix like 'df_', you can identify those by...
dfNames = []
for elem in alldfs:
if str(elem)[:3] == 'df_':
dfNames.append(elem)
... and then organize them in a dict using:
myFrames = {}
for dfName in dfNames:
myFrames[dfName] = eval(dfName)
From that list of interesting dataframes, you can subset those that you'd like to do something with. Here's how you focus only on df_1 and df_2:
invalid = ['df_3']
for inv in invalid:
myFrames.pop(inv, None)
Now you can reference ALL your valid dfs by looping through them:
for key in myFrames.keys():
print(myFrames[key])
And that should cover the...
How do I, in general, access df_1 using an iterator?
...part of the question.
And you can of course reference a single dataframe by its name / key in the dict:
print(myFrames['df_1'])
From here you can do something with ALL columns in ALL dataframes.
for key in myFrames.keys():
myFrames[key] = myFrames[key]*10
print(myFrames[key])
Or, being a bit more pythonic, you can specify a lambda function and apply that to a subset of columns
# A function
decimator = lambda x: x/10
# A subset of columns:
myCols = ['SPEED1', 'SPEED2']
Apply that function to your subset of columns in your dataframes of interest:
for key in myFrames.keys():
for col in list(myFrames[key]):
if col in myCols:
myFrames[key][col] = myFrames[key][col].apply(decimator)
print(myFrames[key][col])
So, back to your function...
modify(df_1,1)
... here's my take on it wrapped in a function.
First we'll redefine the dataframes and the function.
Oh, and with this setup, you're going to have to obtain all dfs OUTSIDE your function with alldfs = [var for var in dir() if isinstance(eval(var), pd.core.frame.DataFrame)].
Here's the datasets and the function for an easy copy-paste:
# Imports
import pandas as pd
import numpy as np
# A few dataframes with random numbers
# df_1
np.random.seed(123)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_1 = pd.DataFrame(np.random.randint(100,150,size=(rows, 3)), columns=['SPEED1', 'SPEED2', 'SPEED3'])
df_1 = df_1.set_index(rng)
# df_2
np.random.seed(456)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_2 = pd.DataFrame(np.random.randint(100,150,size=(rows, 3)), columns=['SPEED1', 'SPEED2', 'SPEED3'])
df_2 = df_2.set_index(rng)
# df_3
np.random.seed(789)
rows = 12
rng = pd.date_range('1/1/2017', periods=rows, freq='D')
df_3 = pd.DataFrame(np.random.randint(100,150,size=(rows, 3)), columns=['SPEED1', 'SPEED2', 'SPEED3'])
df_3 = df_3.set_index(rng)
# A function that divides columns by 10
decimator = lambda x: x/10
# A reference to all available dataframes
alldfs = [var for var in dir() if isinstance(eval(var), pd.core.frame.DataFrame)]
# A function as per your request
def modify(dfs, cols, fx):
""" Define a subset of available dataframes and list of interesting columns, and
apply a function on those columns.
"""
# Subset all dataframes with names that start with df_
dfNames = []
for elem in alldfs:
if str(elem)[:3] == 'df_':
dfNames.append(elem)
# Organize those dfs in a dict if they match the dataframe names of interest
myFrames = {}
for dfName in dfNames:
if dfName in dfs:
myFrames[dfName] = eval(dfName)
print(myFrames)
# Apply fx to the cols of your dfs subset
for key in myFrames.keys():
for col in list(myFrames[key]):
if col in cols:
myFrames[key][col] = myFrames[key][col].apply(decimator)
# A testrun. Results in screenshots below
modify(dfs = ['df_1', 'df_2'], cols = ['SPEED1', 'SPEED2'], fx = decimator)
Here are dataframes df_1 and df_2 before manipulation:
Here are the dataframes after manipulation:
Anyway, this is how I would approach it.
Hope you'll find it useful!
Related
Code snippet
def func(a_val, b_val):
...
return new_df
mydf = mydf.append(existing_df.apply(lambda x: func(x['A'], x['B']), axis=1), ignore_index=True)
As per the code snippet shows, I am trying to use apply to iterate over each row in existing_df and return a new_df that need to be append into mydf eventually, but apply only return a Series object and the new_df is converted into a Series where all columns and rows are being throw into 1 single cell after appending into mydf.
Anyway to allow dataframe.apply to return the original dataframe instead?
Update with sample:
import pandas as pd
existing_df = pd.DataFrame({'router': ['RouterA', 'RouterA', 'RouterB', 'RouterB'], 'vpn': ['vpn1', 'vpn2', 'vpn3', 'vpn4']})
cols = ['router', 'vpn', 'peer']
my_df = pd.DataFrame(columns=cols)
def func(router, vpn):
new_df = pd.DataFrame(columns=cols)
# look for extra information based on router + vpn, and return a dataframe. 1 vpn will return multiple peer result, and the result
# will need to return back to my_df.
return new_df
my_df = my_df.append(existing_df.apply(lambda x: func(x['router'], x['vpn']), axis=1))
and the new_df should look something like this
router vpn peer
RouterA vpn1 10.1.1.1
RouterA vpn1 10.1.1.2
RouterA vpn1 10.1.1.3
and append into my_df, so each router+vpn will return a multiple rows dataframe and return back to my_df.
UPDATE (no apply but iterrows)
Reason:
I found similar question (https://stackoverflow.com/a/45946771/7035448) requiring multiple rows per apply and found this as working and somehow accepted answer which uses pd.apply (https://stackoverflow.com/a/13052373/7035448) does not work for me
import pandas as pd
df = pd.DataFrame([[4, 9]] * 2, columns=['router', 'vpn'])
cols = ['router', 'vpn', 'peer']
my_df = pd.DataFrame(columns=cols)
def func(row):
r1 = row['router']
v1 = row['vpn']
return pd.DataFrame({'router': [r1, r1, r1], 'vpn': [v1, v1, v1], 'peer': ['p1', 'p2', 'p3']})
pd.concat([func(row)
for _, row in df.iterrows()], ignore_index=True)
This can be done with itterrows
Yeah in most use cases I only needed series, but when we need dataframe when there is case like apply needs to return list which is split into columns for example taken from pd.apply, in such cases result_type argument can be helpful.
df = pd.DataFrame([[4, 9]] * 3, columns=['A', 'B'])
df.apply(lambda x: [1, 2], axis=1, result_type='expand')
Look at this picture, it should explain
Above use of argument result_type gives you a dataframe instead of series. Since it is not clear on your func will be done, but from what you described rows are being throw into 1 single cell, this should be the way. I guess?
I am looking into creating a big dataframe (pandas) from several individual frames. The data is organized in MF4-Files and the number of source files varies for each cycle. The goal is to have this process automated.
Creation of Dataframes:
df = (MDF('File1.mf4')).to_dataframe(channels)
df1 = (MDF('File2.mf4')).to_dataframe(channels)
df2 = (MDF('File3.mf4')).to_dataframe(channels)
These Dataframes are then merged:
df = pd.concat([df, df1, df2], axis=0)
How can I do this without dynamically creating variables for df, df1 etc.? Or is there no other way?
I have all filepathes in an Array of the form:
Filepath = ['File1.mf4', 'File2.mf4','File3.mf4',]
Now I am thinking of looping through it and create dynamically the data frames df,df1.df1000.... Any advice here?
Edit here is the full code:
df = (MDF('File1.mf4')).to_dataframe(channels)
df1 = (MDF('File2.mf4')).to_dataframe(channels)
df2 = (MDF('File3.mf4')).to_dataframe(channels)
#The Data has some offset:
x = df.index.max()
df1.index += x
x = df1.index.max()
df2.index += x
#With correct index now the data can be merged
df = pd.concat([df, df1, df2], axis=0)
The way I'm interpreting your question is that you have a predefined list you want. So just:
l = []
for f in [ list ... of ... files ]:
df = load_file(f) # however you load it
l.append(df)
big_df = pd.concat(l)
del l, df, f # if you want to clean it up
You therefore don't need to manually specify variable names for your data sub-sections. If you also want to do checks or column renaming between the various files, you can also just put that into the for-loop (or alternatively, if you want to simplify to a list comprehension, into the load_file function body).
Try this:
df_list = [(MDF(file)).to_dataframe(channels) for file in Filepath]
df = pd.concat(df_list)
When I work with data, very often I will have a bunch of similar objects I want to iterate over to do some processing and store the results.
import pandas as pd
import numpy as np
df1 = pd.DataFrame(np.random.randint(0, 1000, 20))
df2 = pd.DataFrame(np.random.randint(0, 1000, 20))
results = []
for df in [df1, df2]:
tmp_result = df.median() # do some rpocessing
results.append(tmp_result) # append results
The problem I have with this is that it's not clear which dataframe the results correspond to. I thought of using the objects as keys for a dict, but this won't always work as dataframes are not hashable objects and can't be used as keys to dicts:
import pandas as pd
import numpy as np
df1 = pd.DataFrame(np.random.randint(0, 1000, 20))
df2 = pd.DataFrame(np.random.randint(0, 1000, 20))
results = {}
for df in [df1, df2]:
tmp_result = df.median() # do some rpocessing
results[df] = tmp_result # doesn't work
I can think of a few hacks to get around this, like defining unique keys for the input objects before the loop, or storing the input and the result as a tuple in the results list. But in my experience those approaches are rather unwieldy, error prone, and I suspect they're not terrilbly great for memory usage either. Mostly, I just end up using the first example, and make sure I'm careful to manually keep track of the position of the results.
Are there any obvious solutions or best practices to this problem here?
You can keep the original dataframe and the result together in a class:
class Whatever:
def __init__(self, df):
self.df = df
self.result = None
whatever1 = Whatever(pd.DataFrame(...))
whatever2 = Whatever(pd.DataFrame(...))
for whatever in [whatever1, whatever2]:
whatever.result = whatever.df.median()
There are many ways to improve this depending on your situation: generate the result right in the constructor, add a method to generate and store it, compute it on the fly from a property, and so on.
I would concatenate your data frames, adding an index for each data frame, then use a group-by operation.
pd.concat([df1, df2], keys=['df1', 'df2']).groupby(level=0).median()
If your actual processing is more complex, you could use .apply() instead of .median().
You can try something like this:
dd = {'df1': df1,
'df2': df2}
results_dict = {}
for k, v in dd.items():
results_dict[k] = v.mean()
results_df = pd.concat(results_dict, keys=results_dict.keys(), axis=1)
print(results_df)
Output:
df1 df2
0 561.65 549.85
if you want corresponding output dfs , search SO for using globals() in a loop see if you can rename them to something similar to the input.
For df1, you could name it
df1.name = 'df1_output'
then use globals() to set the name of the output df to df1.name. Then you'd have df1 and df1_ouptut
I have multiple dataframes with the same columns but different values that look like that
Product 1 Dataframe
Here's the code that generated them
import pandas as pd
d1 = {"Year":[2018,2019,2020],"Quantity": [10,20,30], "Price": [100,200,300]}
df_product1 = pd.DataFrame(data=d1)
d2 = {"Year":[2018,2019,2020],"Quantity": [20,20,50], "Price": [120,110,380]}
df_product2 = pd.DataFrame(data=d2)
d3 = {"Year":[2018,2019,2020],"Quantity": [40,20,70], "Price": [1000,140,380]}
df_product3 = pd.DataFrame(data=d3)
I merge two dataframes and identify suffixes like so
df_total = df_product1.merge(df_product2,on="Year", suffixes = ("_Product1","_Product2"))
And I get
First Merged Dataframe
However, when I merge another dataframe to the result above using:
df_total = df_total.merge(df_product3,on="Year", suffixes = ("_Product","_Product3"))
I get
Final Merged Dataframe
Where there is no suffix for the third product.
I would like the last two columns of the dataframe to be Quantity_Product3, Price_Product3 instead of just Quantity and Price.
Let me know if it is possible or if I need to approach the problem from a completely different angle.
Why you don't get the result you want
It's explained in the docs: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.merge.html
suffixeslist-like, default is (“_x”, “_y”) A length-2 sequence where
each element is optionally a string indicating the suffix to add to
overlapping column names in left and right respectively. Pass a value
of None instead of a string to indicate that the column name from left
or right should be left as-is, with no suffix. At least one of the
values must not be None.
Suffixes are added to overlapping column names.
See this example - suffixes are added to column b, because both dataframes have a column b, but not to columns a and c, as they are unique and not in common between the two dataframes.
df1 = pd.DataFrame(columns =['a','b'], data = np.random.rand(10,2))
df2 = pd.DataFrame(columns =['b','c'], data = np.random.rand(10,2), index = np.arange(5,15))
# equivalent to an inner join on the indices
out = pd.merge(df1, df2, how ='inner', left_index = True, right_index = True)
A crude solution
Why don't you just rename the columns manually? Not elegant but effective
A possible alternative
The table you are trying to build looks like a pivot. I would look into normalising all your dataframes, concatenating them, then running a pivot on the result.
Depending on your case, this may well be more convoluted and could well be overkill. I mention it because I want to bring your attention to the concepts of pivoting/unpivoting (stacking/unstacking/normalising) data.
The code below takes a df which looks similar to yours and normalises it. For simpler cases you can use pandas.melt(). I don't have the exact data of your example but this should be a good starting point.
def random_dates(start, end, n, unit='D', seed=None):
ndays = (end - start).days + 1
return start + pd.to_timedelta(
np.random.randint(0, ndays, n), unit=unit
)
df = pd.DataFrame()
mysize = 20
df['key'] = np.arange(0,mysize)
df['A_value'] = np.random.randint(0,10000,mysize)
df['A_date'] = random_dates(pd.to_datetime('2010-01-01' ), pd.to_datetime('2019-01-01'), mysize)
df['B_value'] = np.random.randint(-5000,5000,mysize)
df['B_date'] = random_dates(pd.to_datetime('2005-01-01' ), pd.to_datetime('2015-01-01'), mysize)
df['C_value'] = np.random.randint(-10000,10000,mysize)
df['C_date'] = random_dates(pd.to_datetime('2000-01-01' ), pd.to_datetime('2019-01-01'), mysize)
df2 = df.set_index('key', drop=True, verify_integrity = True)
df2 = df2.stack().reset_index()
df2.columns=['key','rawsource','rawvalue']
df2['source'] = df2['rawsource'].apply(lambda x: x[0:1])
df2['metric'] = df2['rawsource'].apply(lambda x: x[2:])
df2 = df2.drop(['rawsource'], axis = 1)
df_piv = df2.pivot_table( index=['key','source'], columns = 'metric' , values ='rawvalue', aggfunc='first' ).reset_index().rename_axis(None, axis=1)
I'm using Pandas library for remote sensing time series analysis. Eventually I would like to save my DataFrame to csv by using chunk-sizes, but I run into a little issue. My code generates 6 NumPy arrays that I convert to Pandas Series. Each of these Series contains a lot of items
>>> prcpSeries.shape
(12626172,)
I would like to add the Series into a Pandas DataFrame (df) so I can save them chunk by chunk to a csv file.
d = {'prcp': pd.Series(prcpSeries),
'tmax': pd.Series(tmaxSeries),
'tmin': pd.Series(tminSeries),
'ndvi': pd.Series(ndviSeries),
'lstm': pd.Series(lstmSeries),
'evtm': pd.Series(evtmSeries)}
df = pd.DataFrame(d)
outFile ='F:/data/output/run1/_'+str(i)+'.out'
df.to_csv(outFile, header = False, chunksize = 1000)
d = None
df = None
But my code get stuck at following line giving a Memory Error
df = pd.DataFrame(d)
Any suggestions? Is it possible to fill the Pandas DataFrame chunk by chunk?
If you know each of these are the same length then you could create the DataFrame directly from the array and then append each column:
df = pd.DataFrame(prcpSeries, columns=['prcp'])
df['tmax'] = tmaxSeries
...
Note: you can also use the to_frame method (which allows you to (optionally) pass a name - which is useful if the Series doesn't have one):
df = prcpSeries.to_frame(name='prcp')
However, if they are variable length then this will lose some data (any arrays which are longer than prcpSeries). An alternative here is to create each as a DataFrame and then perform an outer join (using concat):
df1 = pd.DataFrame(prcpSeries, columns=['prcp'])
df2 = pd.DataFrame(tmaxSeries, columns=['tmax'])
...
df = pd.concat([df1, df2, ...], join='outer', axis=1)
For example:
In [21]: dfA = pd.DataFrame([1,2], columns=['A'])
In [22]: dfB = pd.DataFrame([1], columns=['B'])
In [23]: pd.concat([dfA, dfB], join='outer', axis=1)
Out[23]:
A B
0 1 1
1 2 NaN