I'm trying to learn more about csv maliputation in python and created a csv file. What I'm trying to figure out is how to replace a zero value with a value before it.
1
-
I'd recommend looking into the pandas package for python. Especially for these kinds of tasks there are many many functions that will do it for youMagellan88– Magellan882019-12-13 08:03:59 +00:00Commented Dec 13, 2019 at 8:03
Add a comment
|
4 Answers
Something along the lines of this should work, if you are willing to install pandas
(pip install pandas)
import pandas as pd
D = pd.read_csv("Stock....")
D = D[["desiredColumn"]].replace(0, pd.np.nan )
D = D[["desiredColumn"]].fillna(method="ffill")
In this example all zeros are first replaced with NAN. Once this is done you can use the fillna function that will fill all NAN. This function has a convinience function to either forward fill (ffill) or backward fill (bfill) all NANs
1 Comment
Ajay Shah
The person wants to replace a zero value with the value before it. That's the question. Isn't it?
This is not something specific to csv's. You can do it easily as below.
import csv
with open("input.csv") as file:
reader = csv.reader(file)
for row in reader:
res = [] # load this list with the sanitized `values`
for v in row:
res.append(v if float(v) else res[-1]) # important! first element of `row` must be non-zero
print(res) # use `res` instead of `row`
As noted, this assumes each row starts with a non-zero value.
If this is not guaranteed, a simple solution would be initializing res list with a fallback value for starting zeros then discard the first element of the res once it is loaded.
I.e.
import csv
with open("input.csv") as file:
reader = csv.reader(file)
for row in reader:
res = [99] # if first value of `row` is zero, consider it as `99`
for v in row:
res.append(v if float(v) else res[-1]) # important! first element of r must be non-zero
res = res[1:] # discard the first element after `res` is loaded
print(res)
Comments
If you want to do it in csv module no matter what, then you can do it by just adding a few lines to your if-else statement. Just add:
for row in range(len(reader)):
for i in range(len(reader[row])):
if (reader[row][i] == '0'):
if i: #Check if there's a number before the occurrence of 0
reader[row][i] = reader[row][i-1]
else:
pass
Otherwise, if you aren't really particular with your module, then try using pandas, which is a much more powerful package than the csv package.
3 Comments
Alex
I was going to answer the same; you should also add a check that we're not already on the first element, because if i == 0, we'd end at [-1], and that's not what we want...
Seraph Wedd
I know, that's what the
if i: statement is. If i is anything other than zero, it will return True.Alex
ops, I was looking for it... and still missed it, sorry!
You should transfer the data to its own data structure. The default choice would be a list, unless you want to use it with a library that needs something else input.
import csv
values = list()
with open('file.csv') as file:
reader = csv.reader(file)
for row in reader:
value = int(row[1])
if value == 0:
if values:
values.append( values[-1] )
else:
# Whatever is meaningful in this case
pass
else:
values.append( value )
# do stuff with values here
PS: You make answerers' lives easier if you paste your code as code and not as a screenshot of code. This might make the difference of getting an answer or not in one of your future questions.
1 Comment
Anubis
int(row[1]) will crash if row[1] is a string representing a floating point. e.g. 1.234