rp 0.1.981__tar.gz → 0.1.983__tar.gz

{rp-0.1.981 → rp-0.1.983}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 1.0
 Name: rp
-Version: 0.1.981
+Version: 0.1.983
 Summary: Ryan's Python
 Home-page: https://github.com/RyannDaGreat/Quick-Python
 Author: Ryan Burgert

{rp-0.1.981 → rp-0.1.983}/rp/r.py

@@ -1382,7 +1382,7 @@ def uniform_float_color_image(height:int,width:int,color:tuple):
     #    image=tiled_images(random_color_tiles,border_thickness=0)
     #    display_image(image) #The result will look like https://i.imgur.com/COlmGRT.png 
     #
-    assert height>0 and width>0
+    assert height>=0 and width>=0
     assert is_number(color) or is_color(color) and len(color) in {3,4}, 'Color should be a number, an RGB float color, or an RGBA float color'
     
     if is_number(color):
@@ -2212,6 +2212,21 @@ def random_batch(full_list,batch_size: int = None,*,retain_order: bool = False):
     #       list a ≣ ⨀ a None True
     #       b ≣ len ⨀ a b
 
+    # Check if the input is a pandas DataFrame by class name and presence of .iloc
+    if full_list.__class__.__name__ == 'DataFrame' and hasattr(full_list, 'iloc'):
+        if batch_size is None:
+            batch_size = len(full_list)
+        assert 0 <= batch_size <= len(full_list), "batch_size must be between 0 and the number of rows in the DataFrame"
+        
+        # Use random_batch recursively to select random row indices
+        random_indices = random_batch(
+            range(len(full_list)),
+            batch_size,
+            retain_order=retain_order,
+        )
+
+        return full_list.iloc[random_indices]
+
     from collections.abc import Mapping, Set, Iterable
     if isinstance(full_list, Mapping):
         #Make it work with dicts too
@@ -2222,10 +2237,11 @@ def random_batch(full_list,batch_size: int = None,*,retain_order: bool = False):
 
     if isinstance(full_list,set):
         full_list=list(full_list)
+
     if batch_size is None:  # The default if not specified
         # If we don't specify the batch size, assume that we simply want a shuffled version of the full_list
         if retain_order:
-            return full_list  # A result of the rCode algebra. This simply speeds up the process.
+            return list(full_list)  # A result of the rCode algebra. This simply speeds up the process.
         batch_size=len(full_list)
     else:
         assert 0 <= batch_size <= len(full_list),"batch_size == " + str(batch_size) + " ⋀ len(full_list) == " + str(len(full_list)) + "，∴  ¬ (0 <= batch_size <= len﹙full_list﹚)   Explanation: We do not allow duplicates, ∴ we cannot generate a larger batch than we have elements to choose from full_list"
@@ -2243,11 +2259,14 @@ def random_batch_up_to(full_list, max_batch_size=None, retain_order=False):
     Like random batch, but when batch_size is larger than the full_list, the output will only have the length of the input full list
     This behaviour is as opposed to returning an output with the length of batch_size when len(full_list)<batch_size
     """
-    if batch_size is not None:
-        assert isinstance(batch_size, int)
-        batch_size = min(len(full_list), batch_size)
 
-    return random_batch(full_list, max_batch_size, retain_order = retain_order)
+    if max_batch_size is not None:
+        assert isinstance(max_batch_size, int)
+        batch_size = min(len(full_list), max_batch_size)
+    else:
+        batch_size = max_batch_size
+
+    return random_batch(full_list, batch_size, retain_order = retain_order)
 
 def random_substring(string:str,length:int):
     assert len(string)>=length
@@ -3129,6 +3148,7 @@ def save_image_to_imgur(image):
                 
 def save_image_jpg(image,path,*,quality=100,add_extension=True):
     #If add_extension is True, will add a '.jpg' or '.jpeg' extension to path IFF it doesn't allready end with such an extension (AKA 'a/b/c.jpg' -> 'a/b/c.jpg' BUT 'a/b/c.png' -> 'a/b/c.png.jpg')
+    make_directory(get_parent_folder(path)) #Make sure the directory exists
     image=np.asarray(image)
     image=as_rgb_image(image)
     image=as_byte_image(image)
@@ -6375,6 +6395,7 @@ def save_pickled_value(file_name: str,*variables):
 # endregion
 # region  .txt ⟷ str: ［string_to_text_file，text_file_to_string］
 def string_to_text_file(file_path: str,string: str,) -> None:
+    "string_to_text_file(file_path, string) writes text file"
     file_path=get_absolute_path(file_path)#Make sure it recognizes ~/.vimrc AKA with the ~ attached
     file=open(file_path,"w")
     try:
@@ -6389,6 +6410,7 @@ def string_to_text_file(file_path: str,string: str,) -> None:
 
 _text_file_to_string_cache={}
 def text_file_to_string(file_path: str,use_cache=False) -> str:
+    "text_file_to_string(file_path) reads text file"
     #Only reason not to use use_cache is if you're worried about memory consumption
     #   It will refresh the cache entry if it's out of date even if use_cache is True.
     #   TODO: Make this happen on load_image, etc...all other functions that read from a file and have use_cache as an option
@@ -12045,14 +12067,20 @@ def pseudo_terminal(*dicts,get_user_input=python_input,modifier=None,style=pseud
         S   $os.system("sh");
         Z   $os.system("zsh");
 
+        JL PYM jupyter lab
+        UNCOMMIT !git reset --soft HEAD^
+
         NB  $extract_code_from_ipynb()
         NBA  $extract_code_from_ipynb(ans)
         NBC  $r._clear_jupyter_notebook_outputs()
         NBCA $r._nbca(ans) # Clear a notebook
         NBCH $r._nbca($get_all_files(file_extension_filter='ipynb')) #Clear all notebooks in the current directory
         NBCHY $r._nbca($get_all_files(file_extension_filter='ipynb',sort_by='size')[::-1], auto_yes=True) #Clear all notebooks in the current directory without confirmation
+        NBCHYF $r._nbca($get_all_files(file_extension_filter='ipynb',sort_by='size')[::-1], auto_yes=True,parallel=True) #Clear all notebooks in the current directory without confirmation
         NCA  $r._nbca(ans)
 
+        JL PYM jupyter lab
+
         IPYK $add_ipython_kernel()
 
         INS $input_select("Select:", $line_split(ans) if isinstance(ans,str) else ans)
@@ -17668,8 +17696,16 @@ def labeled_image(image,
 
 def labeled_images(images,labels,*args,**kwargs):
     #The plural of labeled_image
+    assert is_iterable(labels)
+    assert is_iterable(images)
+
     images=list(images)
-    labels=list(labels)
+
+    if isinstance(labels,str):
+        labels=[labels]*len(images)
+    else:    
+        labels=list(labels)
+
     assert len(images)==len(labels)
     return [labeled_image(image,label,*args,**kwargs) for image,label in zip(images,labels)]
 
@@ -18580,6 +18616,13 @@ def horizontally_concatenated_images(*image_list):
     #This is different from np.column_stack because it handles images of different resolutions.
     #It also can mix RGB, greyscale, and RGBA images.
 
+    #try:
+    #    #Might be faster than converting all to RGBA then doing height stuff to them all
+    #    print("HORCAT SHORTCUT")
+    #    return np.concatenate(as_numpy_array(image_list), axis=2)
+    #except Exception as e:
+    #    pass
+
     image_list=[as_rgba_image(as_float_image(image)) for image in image_list]#Right now, bring the images to the max possible format for compatiability. Might make it smarter later such as to preserve the format if all input images are binary, for example.
     max_height=max(img.shape[0]for img in image_list)
     def heightify(img):
@@ -18646,8 +18689,8 @@ def grid_concatenated_images(image_grid):
     image_grid=list(image_grid)
     max_image_widths=[0]*max(map(len,image_grid))
 
-    #A 1x1 transparent pixel
-    null_img = uniform_float_color_image(1,1,(0,0,0,0))
+    #A 0x0 transparent pixel
+    null_img = uniform_float_color_image(0,0,(0,0,0,0))
 
     for y,image_row in enumerate(image_grid):
         image_grid[y]=image_row=list(image_row)
@@ -19437,6 +19480,10 @@ def bordered_image_solid_color(image,color=(1.,1.,1.,1.),thickness=1,width=None,
     
     return image
 
+def bordered_images_solid_color(*images,color=(1.,1.,1.,1.),thickness=1,width=None,height=None,top=None,bottom=None,left=None,right=None):
+    images=detuple(images)
+    return [bordered_image_solid_color(x, color, thickness, width, height, top, bottom, left, right) for x in images]
+
 def get_principle_components(tensors,number_of_components=None):
     #Returns orthogonal, normalized, sorted-by-eigenvalue-in-descending-order principle components (retaining the shape of the original tensors, to make eigenfaces easy to extract for example)
     #For example, if we feed get_principle_components a list of images, we expect to return a list of images (not a list of vectors, like most PCA implementations require). This is for the sake of convenience.
@@ -20980,11 +21027,12 @@ def crop_image(image, height: int = None, width: int = None, origin='top left'):
     
     return out
 
-def crop_image_zeros(image):
+def crop_image_zeros(image,*,output='image'):
     """
     #Given some big image that is surrounded by black, or 0-alpha transparency, crop out that excess region
     #TODO: Give examples
     """
+    assert output in 'image bounds'.split(),'output is a string indicating what the output type is - it can be either image or bounds but you gave type '+str(type(output))+' '+str(output)
     assert is_image(image),'Error: input is not an image as defined by rp.is_image()'
     if is_grayscale_image(image):
         points=np.argwhere(image)#Crop out the black regions
@@ -20998,9 +21046,11 @@ def crop_image_zeros(image):
         points=[[0,0],[0,0]]
     top,left=np.min(points,axis=0)
     bottom,right=np.max(points,axis=0)+1
-    cropped=image[top:bottom,left:right]
-    #print(top,bottom,left,right)
-    return cropped
+    if output=='image':
+        cropped=image[top:bottom,left:right]
+        return cropped
+    if output=='bounds':
+        return top,bottom,left,right
 
 def cv_contour_to_segment(contour):
     """
@@ -26610,14 +26660,22 @@ def _display_filetype_size_histogram(root='.'):
     _maybe_display_string_in_pager(printed_lines)
     print(printed_lines)
 
-def _nbca(paths,auto_yes=False):
+def _nbca(paths,auto_yes=False,parallel=False):
     if isinstance(paths,str):
         paths=line_split(paths)
-    for x in paths:
-        if file_exists(x) and x.endswith('.ipynb'):
-            _clear_jupyter_notebook_outputs(x,auto_yes=auto_yes)
-        else:
-            print(fansi("r.nbca: Skipping; not a ipynb file:",'yellow'),x)
+    def do_path(x):
+        try:
+            if file_exists(x) and x.endswith('.ipynb'):
+                _clear_jupyter_notebook_outputs(x,auto_yes=auto_yes)
+            else:
+                print(fansi("r.nbca: Skipping; not a ipynb file:",'yellow'),x)
+        except Exception as e:
+            pass
+    if not parallel:
+        for x in paths:
+            do_path(x)
+    else:
+        par_map(do_path,paths)
 
 def _clear_jupyter_notebook_outputs(path:str=None, auto_yes=False):
     #This clears all outputs of a jupyter notebook file
@@ -26960,7 +27018,7 @@ def as_numpy_images(images):
     else:
         if not is_iterable(images):
             raise TypeError('Unsupported image datatype: %s ; its not even iterable!'%type(images))
-        if all(is_pil_image(x) for x in images):
+        if all((is_pil_image(x) or _is_torch_tensor(x) or _is_numpy_array(x)) for x in images):
             return [as_numpy_image(x) for x in images]
         else:
             raise TypeError('Unsupported image datatype: %s of %s'%(type(images),repr(set(map(type,images)))))

rp-0.1.983/rp/version.py

@@ -0,0 +1 @@
+983

rp-0.1.983/rp.egg-info/PKG-INFO

@@ -0,0 +1,11 @@
+Metadata-Version: 1.0
+Name: rp
+Version: 0.1.983
+Summary: Ryan's Python
+Home-page: https://github.com/RyannDaGreat/Quick-Python
+Author: Ryan Burgert
+Author-email: ryancentralorg@gmail.com
+License: UNKNOWN
+Description: UNKNOWN
+Keywords: not_searchable_yet_go_away_until_later_when_this_is_polished
+Platform: UNKNOWN

rp-0.1.983/version.py

@@ -0,0 +1 @@
+983

rp-0.1.981/rp/version.py

@@ -1 +0,0 @@
-981

rp-0.1.981/rp.egg-info/PKG-INFO

@@ -1,11 +0,0 @@
-Metadata-Version: 1.0
-Name: rp
-Version: 0.1.981
-Summary: Ryan's Python
-Home-page: https://github.com/RyannDaGreat/Quick-Python
-Author: Ryan Burgert
-Author-email: ryancentralorg@gmail.com
-License: UNKNOWN
-Description: UNKNOWN
-Keywords: not_searchable_yet_go_away_until_later_when_this_is_polished
-Platform: UNKNOWN

rp-0.1.981/version.py

@@ -1 +0,0 @@
-981