Print numpy array without ellipsis

Question

I want to print a numpy array without truncation. I have seen other solutions but those don't seem to work.

Here is the code snippet:

total_list = np.array(total_list)
np.set_printoptions(threshold=np.inf)
print(total_list)

And this is what the output looks like:

22        A
23        G
24        C
25        T
26        A
27        A
28        A
29        G
         ..
232272    G
232273    T
232274    G
232275    C
232276    T
232277    C
232278    G
232279    T

This is the entire code. I might be making a mistake in type casting.

import csv
import pandas as pd
import numpy as np



seqs = pd.read_csv('BAP_GBS_BTXv2_imp801.hmp.csv')
plts = pd.read_csv('BAP16_PlotPlan.csv')

required_rows = np.array([7,11,14,19,22,31,35,47,50,55,58,63,66,72,74,79,82,87,90,93,99])
total_list = []


for i in range(len(required_rows)):
    curr_row = required_rows[i];
    print(curr_row)
    for j in range(len(plts.RW)):
        if(curr_row == plts.RW[j]):
            curr_plt = plts.PI[j]
            curr_range = plts.RA1[j]
            curr_plt = curr_plt.replace("_", "").lower()
            if curr_plt in seqs.columns:
                new_item = [curr_row,curr_range,seqs[curr_plt]]
                total_list.append(new_item)
                print(seqs[curr_plt]) 


total_list = np.array(total_list)
'''
np.savetxt("foo.csv", total_list[:,2], delimiter=',',fmt='%s')
total_list[:,2].tofile('seqs.csv',sep=',',format='%s')
'''
np.set_printoptions(threshold='nan')

print(total_list)

It works for me, maybe your total_list is not a numpy array? — Psidom, Jun 01 '17 at 15:33
It is not how numpy arrays look. Are you sure it is not a pandas Series? — ayhan, Jun 01 '17 at 15:34
That was the actual intent. But had the same problem. So I m trying to print it on the terminal first. — Harjatin, Jun 01 '17 at 15:43
does this file need to be human readable? or is it for the purpose of storing the data? — Aaron, Jun 01 '17 at 15:51
Numpy has built in functions to do either... [`np.savetxt`](https://docs.scipy.org/doc/numpy/reference/generated/numpy.savetxt.html) can be configured to write a standard csv file, and [`np.save`](https://docs.scipy.org/doc/numpy/reference/generated/numpy.save.html#numpy.save) can save to a numpy specific binary format. — Aaron, Jun 01 '17 at 15:56
Please check the edit. I have tried that. It also saves the csv in truncated format. — Harjatin, Jun 01 '17 at 15:58

score 32 · Answer 1 · edited Jan 04 '20 at 12:55

32

use the following snippet to get no ellipsis.

import numpy
import sys
numpy.set_printoptions(threshold=sys.maxsize)

EDIT:

If you have a pandas.DataFrame use the following snippet to print your array:

def print_full(x):
    pd.set_option('display.max_rows', len(x))
    print(x)
    pd.reset_option('display.max_rows')

Or you can use the pandas.DataFrame.to_string() method to get the desired result.

EDIT':

An earlier version of this post suggested the option below

numpy.set_printoptions(threshold='nan')

Technically, this might work, however, the numpy documentation specifies int and None as allowed types. Reference: https://docs.scipy.org/doc/numpy/reference/generated/numpy.set_printoptions.html.

edited Jan 04 '20 at 12:55

Nico Schlömer

53,797
27
201
249

answered Jun 01 '17 at 15:35

Szabolcs Dombi

5,493
3
39
71

python and numpy version? – Szabolcs Dombi Jun 01 '17 at 15:42
Python 3.5.2, numpy 1.11.1 – Harjatin Jun 01 '17 at 16:01
@Harjatin maybe the other print is not formatted properly (the first one in the loop) – Szabolcs Dombi Jun 01 '17 at 16:05
I commented all the other print statements. The output is still truncated. – Harjatin Jun 01 '17 at 16:10
@Harjatin are you sure it is not a pandas.DataFrame? read [this](https://stackoverflow.com/questions/19124601/is-there-a-way-to-pretty-print-the-entire-pandas-series-dataframe) – Szabolcs Dombi Jun 01 '17 at 16:16
1

At least from 1.16.5, this doesn't work: `ValueError: threshold must be numeric and non-NAN, try sys.maxsize for untruncated representation`, so use `sys.maxsize`. – Jim Hunziker Dec 13 '19 at 20:22

a p · Answer 2 · 2017-06-01T15:54:07.643

5

You can get around the weird Numpy repr/print behavior by changing it to a list:

print list(total_list)

should print out your list of 2-element np arrays.

edited Jun 01 '17 at 15:54

answered Jun 01 '17 at 15:39

a p

3,098
2
24
46

This still gives the same output – Harjatin Jun 01 '17 at 15:42
2

Weird - does it happen if you use `total_list.tolist()` as well? – a p Jun 01 '17 at 15:45
1

you probably run a different code what you edit :)) I can confirm both of the answers work fine. add a `print("hello")` or whatever – Szabolcs Dombi Jun 01 '17 at 15:45
What is `total_list` to start with? – a p Jun 01 '17 at 15:52
numpy's printing is more useful though for most cases, because it is more readable ― no commas between the numbers.. – matanster Sep 02 '18 at 17:40

score 4 · Answer 3 · answered Jun 01 '17 at 16:21

4

You are not printing numpy arrays.

Add the following line after the imports:

pd.set_option('display.max_rows', 100000)

answered Jun 01 '17 at 16:21

Szabolcs Dombi

5,493
3
39
71

Thank you. This worked. I am little confused through, shouldn't `total_list = np.array(total_list)` covert DataFrame to numpy array ? – Harjatin Jun 01 '17 at 16:49

score 1 · Answer 4 · answered Dec 16 '18 at 14:50

#for a 2d array
def print_full(x):
    dim = x.shape
    pd.set_option('display.max_rows', dim[0])#dim[0] = len(x)
    pd.set_option('display.max_columns', dim[1])
    print(x)
    pd.reset_option('display.max_rows')
    pd.reset_option('display.max_columns')

score 1 · Answer 5 · answered Sep 11 '19 at 14:51

1

It appears that as of Python 3, the threshold can no longer be unlimited.

Therefore, the recommended option is:

import numpy
import sys
numpy.set_printoptions(threshold=sys.maxsize)

answered Sep 11 '19 at 14:51

simonalexander2005

4,338
4
48
92

Print numpy array without ellipsis

5 Answers5

Linked