TypeError in pyvips

[Rasaa84]

Hi,

I am going to use pyvips to create dzi files from a numpy array but get the following error.

TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_56220/2408616579.py in <module>
      3 
      4 a = (np.random.random((100, 100, 3)) * 255).astype(np.uint8)
----> 5 image = pyvips.Image.new_from_array(a)

~/anaconda3/lib/python3.7/site-packages/pyvips/vimage.py in new_from_array(array, scale, offset)
    301         for y in range(0, height):
    302             for x in range(0, width):
--> 303                 a[x + y * width] = array[y][x]
    304 
    305         vi = vips_lib.vips_image_new_matrix_from_array(width, height, a, n)

TypeError: only size-1 arrays can be converted to Python scalars

I am using Ubuntu 18.04.6. Thanks in advance for your help.

Best,

[jcupitt]

Hi @Rasaa84,

It seems to be working for me. I see:

$ python3
Python 3.10.7 (main, Nov 24 2022, 19:45:47) [GCC 12.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> import pyvips
>>> a = (np.random.random((100, 100, 3)) * 255).astype(np.uint8)
>>> a
array([[[147, 168,  35],
        [235, 102,   5],
        [149,   8, 189],
        ...,
        [ 66,   7, 253],
        [ 75, 161, 124],
        [164, 197,  41]],

       [[184, 198, 163],
        [ 81, 113,  50],
        [  4,  82, 160],
        ...,
        [152, 177, 119],
        [123, 129,  59],
        [174, 244, 138]],

       [[104, 243,  90],
        [ 89, 153, 208],
        [ 90,  44, 232],
        ...,
        [176,  46, 127],
        [233, 162, 177],
        [ 91,   8, 140]],

       ...,

       [[118,  10,  11],
        [123, 136, 150],
        [ 50, 138, 165],
        ...,
        [109,   8,  56],
        [163, 144,  48],
        [217,  82, 231]],

       [[ 37, 115, 231],
        [132, 116, 196],
        [ 29, 233, 243],
        ...,
        [216,  58, 221],
        [216, 234, 186],
        [ 22, 121,  50]],

       [[181, 251, 145],
        [ 63,  36, 249],
        [175, 119, 111],
        ...,
        [ 33,   8,  46],
        [ 75, 138, 119],
        [ 31, 120, 114]]], dtype=uint8)
>>> x = pyvips.Image.new_from_array(a)
>>> x
<pyvips.Image 100x100 uchar, 3 bands, srgb>
>>> pyvips.__version__
#'2.2.1'

I would guess this is an anaconda issue. Have you tried vanilla python? Anaconda isn't usually necessary on linux, and often causes bugs.

[jcupitt]

I tried on 18.04 with vanilla python and it seems to work there too. I used this dockerfile:

FROM ubuntu:18.04

ENV DEBIAN_FRONTEND=noninteractive

RUN apt-get update

RUN apt-get install -y \
        build-essential \
        python3-pip

RUN apt-get install -y --no-install-recommends \
        libvips-dev

RUN pip3 install pyvips numpy

Then:

$ docker build -t pyvips-ubuntu18.04 .
Sending build context to Docker daemon  16.38kB
Step 1/6 : FROM ubuntu:18.04
 ---> e28a50f651f9
...
$ docker run -it --rm -v $PWD:/data pyvips-ubuntu18.04
root@7000341907eb:/# python3
Python 3.6.9 (default, Nov 25 2022, 14:10:45) 
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pyvips
>>> pyvips.API_mode
True
>>> import numpy as np
>>> np.__version__
'1.19.5'
>>> a = (np.random.random((100, 100, 3)) * 255).astype(np.uint8)
>>> a
array([[[215, 243,  83],
        [ 54, 163, 127],
        [180,  76, 119],
        ...,
        [246, 202, 209],
        [ 80,  48, 242],
        [137,  37, 125]]], dtype=uint8)
>>> x = pyvips.Image.new_from_array(a)
>>> x
<pyvips.Image 100x100 uchar, 3 bands, rgb>
>>>