Skip to content

Remove usage of fastbook #38

@jfthuong

Description

@jfthuong

Is your feature request related to a problem? Please describe.
Fastbook is a bit overkill for very few features and installation is sometimes buggy on some platforms (e.g. Kaggle AFAIK)

Describe the solution you'd like
Moving similar features to unpackai package

  • Setting contexts (see below) => could be in unpackai/__init__.py
from fastai.vision.all import *
# from fastai.torch_core import set_seed
# import torch

mpl.rcParams['savefig.dpi']= 200
mpl.rcParams['font.size']=12

set_seed(42)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
pd.set_option('display.max_columns',999)
np.set_printoptions(linewidth=200)
torch.set_printoptions(linewidth=200)
  • Function setup_colab (it just mounts Google Drive in Colab) (related to issue Simplify usage of Google Drive #36)
  • Function setup_book (useless => it just setup colab if we are using Colab)
  • Function gv
def gv(s): return graphviz.Source('digraph G{ rankdir="LR"' + s + '; }')
  • Function get_image_files_sorted
def get_image_files_sorted(path, recurse=True, folders=None):
    return get_image_files(path, recurse, folders).sorted()
  • Function search_images_bing (search images with Bing => we recommend DDG so don't need it)
  • Function search_images_ddg (we use another utility => don't need it)
  • Function plot_function
  • Function draw_tree
  • Function cluster_columns (dendogram)
  • Function image_cat (returns BytesIO object from an image of cat) => probably don't need it
  • Function image_bear (same but for a bear) => probably don't need it

Here is the part of the code copied from __init__.py of fastbook:

__version__ = "0.0.18"
import matplotlib as mpl, pkgutil, requests, time
from fastai.vision.all import *
from fastdownload import download_url
from pandas.api.types import CategoricalDtype
from scipy.cluster import hierarchy as hc
from io import StringIO, BytesIO
from urllib.error import URLError,HTTPError

try: from ipywidgets import widgets
except ModuleNotFoundError: warn("Missing `ipywidgets` - please install it")
try: import sentencepiece
except ModuleNotFoundError: warn("Missing `sentencepiece` - please run `pip install 'sentencepiece<0.1.90'`")
try: from nbdev.showdoc import *
except ModuleNotFoundError: warn("Missing `nbdev` - please install it")
try:
    import graphviz
    from sklearn.tree import export_graphviz
except ModuleNotFoundError: warn("Missing `graphviz` - please run `conda install fastbook`")

[...]

def plot_function(f, tx=None, ty=None, title=None, min=-2, max=2, figsize=(6,4)):
    x = torch.linspace(min,max)
    fig,ax = plt.subplots(figsize=figsize)
    ax.plot(x,f(x))
    if tx is not None: ax.set_xlabel(tx)
    if ty is not None: ax.set_ylabel(ty)
    if title is not None: ax.set_title(title)

def draw_tree(t, df, size=10, ratio=0.6, precision=0, **kwargs):
    s=export_graphviz(t, out_file=None, feature_names=df.columns, filled=True, rounded=True,
                      special_characters=True, rotate=False, precision=precision, **kwargs)
    return graphviz.Source(re.sub('Tree {', f'Tree {{ size={size}; ratio={ratio}', s))

def cluster_columns(df, figsize=(10,6), font_size=12):
    corr = np.round(scipy.stats.spearmanr(df).correlation, 4)
    corr_condensed = hc.distance.squareform(1-corr)
    z = hc.linkage(corr_condensed, method='average')
    fig = plt.figure(figsize=figsize)
    hc.dendrogram(z, labels=df.columns, orientation='left', leaf_font_size=font_size)
    plt.show()

Metadata

Metadata

Labels

enhancementNew feature or request

Type

No type

Projects

No projects

Relationships

None yet

Development

No branches or pull requests

Issue actions