ivsanro1

import streamlit as st

if 'show_a' not in st.session_state:
    st.session_state.show_a = True

if 'show_b' not in st.session_state:
    st.session_state.show_b = False

st.header("Alternating Buttons")

ivsanro1

ivsanro1

Description

Long text, code snippet, image, etc.

ivsanro1

Disclaimer

BEWARE OF USING IT WITH CACHED FUNCTIONS!

Unexpected behavior when used with cached functions. Disable all cache (like functools.lru_cache) that don't let the functions run normally in certain conditions.

THIS IS FOR ACADEMIC/LEARNING PURPOSES!

Do not blindly rely on the results this can give you as it has not been fully tested.

Usage example

ivsanro1

import pandas as pd

def json_normalize_and_concat(df: pd.DataFrame, col: str) -> pd.DataFrame:
    index_name = df.reset_index().columns[0]
    
    df_json_norm = pd.json_normalize(df[col], max_level=0).reset_index(drop=True)
    df_ret = pd.concat([
        df.reset_index().drop([c for c in df_json_norm.columns if c in df.columns], axis=1),
        df_json_norm
    ], axis=1).set_index(index_name).drop(col, axis=1)

ivsanro1

from thefuzz import fuzz
from scipy.optimize import linear_sum_assignment
    
def df_merge_fuzzy_optimal_with_cost(
    df_left: pd.DataFrame,
    df_right: pd.DataFrame,
    left_on: str,
    right_on: str,
    fn_cost: Optional[Callable] = lambda x, y: 100 - fuzz.ratio(x, y)
) -> pd.DataFrame:

ivsanro1

import git
import os
from pathlib import Path

def get_root() -> Path:
    return Path(git.Repo(os.getcwd(), search_parent_directories=True).working_dir)

ivsanro1

ivsanro1

def read_avro(bytes_buff_or_file):
    import avro.schema
    from avro.datafile import DataFileReader, DataFileWriter
    from avro.io import DatumReader, DatumWriter
    from io import BytesIO
    import json
    if type(bytes_buff_or_file) == bytes:
        from tempfile import TemporaryFile
        # Has to be a file yes or yes because of how avro lib is made
        with TemporaryFile() as tmp:

ivsanro1

from tensorflow.keras.applications.resnet50 import ResNet50, preprocess_input, decode_predictions
from tensorflow.keras.preprocessing import image as tf_image_preproc

model = ResNet50(weights='imagenet')

from tensorflow.keras.models import Model

resnet50_without_softmax = Model(inputs=model.input,
                                 outputs=model.get_layer('avg_pool').output)