voice keyword blacklist

This commit is contained in:
2025-12-17 01:19:22 -08:00
parent eee6947f01
commit 417273c745

View File

@@ -22,18 +22,22 @@ def _():
tqdm.pandas()
TAGUETTE_EXPORT_DIR = Path('./data/processing/02_taguette_export')
WORKING_DIR = Path('./data/processing/02-b_WordClouds')
VOICE_EXCLUDE_KEYWORDS_FILE = WORKING_DIR / 'voice_excl_keywords.txt'
if not WORKING_DIR.exists():
WORKING_DIR.mkdir(parents=True)
if not TAGUETTE_EXPORT_DIR.exists():
TAGUETTE_EXPORT_DIR.mkdir(parents=True)
if not VOICE_EXCLUDE_KEYWORDS_FILE.exists():
VOICE_EXCLUDE_KEYWORDS_FILE.touch()
return (
OLLAMA_LOCATION,
TAGUETTE_EXPORT_DIR,
VOICE_EXCLUDE_KEYWORDS_FILE,
WORKING_DIR,
connect_qumo_ollama,
mo,
@@ -115,7 +119,7 @@ def _(all_tags_df, mo):
return (tag_select,)
@app.cell
@app.cell(hide_code=True)
def _(WORKING_DIR, all_tags_df, mo, tag_select):
mo.stop(not tag_select.value, mo.md("Select tag to continue"))
@@ -152,7 +156,7 @@ def _(WORKING_DIR, all_tags_df, mo, tag_select):
@app.cell(hide_code=True)
def _(KEYWORD_FREQ_FPATH, mo):
mo.md(rf"""
# 4) Keyword extraction {'(skippable, see 4b)' if KEYWORD_FREQ_FPATH.exists() else ''}
# 4) Keyword extraction {'(skippable, see 4b)' if KEYWORD_FREQ_FPATH.exists() else '(Required)'}
""")
return
@@ -267,14 +271,21 @@ def _(KEYWORD_FREQ_FPATH, mo, start_processing_btn):
load_existing_btn = None
if KEYWORD_FREQ_FPATH.exists():
load_existing_btn = mo.ui.run_button(label=f"Load keywords from `{KEYWORD_FREQ_FPATH.name}`", kind='warn')
load_existing_btn = mo.ui.run_button(label=f"Load `{KEYWORD_FREQ_FPATH.name}`", kind='warn')
load_existing_btn
return (load_existing_btn,)
@app.cell(hide_code=True)
def _(KEYWORD_FREQ_FPATH, freq_df, load_existing_btn, pd):
def _(
KEYWORD_FREQ_FPATH,
VOICE_EXCLUDE_KEYWORDS_FILE,
freq_df,
load_existing_btn,
pd,
tag_select,
):
if load_existing_btn is not None and load_existing_btn.value:
_fdf = pd.read_excel(KEYWORD_FREQ_FPATH, engine='openpyxl')
@@ -284,6 +295,19 @@ def _(KEYWORD_FREQ_FPATH, freq_df, load_existing_btn, pd):
_fdf.reset_index(drop=True, inplace=True)
print(f"Loaded `{KEYWORD_FREQ_FPATH}` successfully.")
if tag_select.value.startswith('V'):
# Read exclusion list
excl_kw = []
with VOICE_EXCLUDE_KEYWORDS_FILE.open('r') as _f:
for line in _f:
excl_kw.append(line.strip())
_drop_idx = _fdf[_fdf['keyword'].isin(excl_kw)].index
_fdf.drop(index=_drop_idx, inplace=True, axis=0)
print(f"Dropped {len(_drop_idx)} keywords automatically")
frequency_df = _fdf
else:
@@ -374,7 +398,15 @@ def _(mo, table_selection):
@app.cell(hide_code=True)
def _(KEYWORD_FREQ_FPATH, frequency_df, mo, remove_rows_btn, table_selection):
def _(
KEYWORD_FREQ_FPATH,
VOICE_EXCLUDE_KEYWORDS_FILE,
frequency_df,
mo,
remove_rows_btn,
table_selection,
tag_select,
):
_s = None
if remove_rows_btn is not None and remove_rows_btn.value:
# get selected rows
@@ -382,7 +414,20 @@ def _(KEYWORD_FREQ_FPATH, frequency_df, mo, remove_rows_btn, table_selection):
if len(selected_rows) >0 :
rows_to_drop = table_selection.value.index.tolist()
try:
if tag_select.value.startswith('V'):
# append values to an VoiceKeywordsExclusion file (txt file just a list of keywords)
exclude_keywords = frequency_df.loc[rows_to_drop, 'keyword'].to_list()
with VOICE_EXCLUDE_KEYWORDS_FILE.open('w') as f:
for _kw in exclude_keywords:
f.write(_kw + '\n')
frequency_df.drop(index=rows_to_drop, inplace=True, axis=0)
except KeyError:
_s = mo.callout("GO BACK TO STEP 4b) and reload data to continue refining the dataset.", kind='warn')
else:
@@ -395,7 +440,7 @@ def _(KEYWORD_FREQ_FPATH, frequency_df, mo, remove_rows_btn, table_selection):
print(f"Updated keyword frequencies saved to: `{KEYWORD_FREQ_FPATH}`")
# mo.callout(f"Updated keyword frequencies saved to: `{KEYWORD_FREQ_FPATH}`", kind="success")
_s = mo.callout("GO BACK TO STEP 4b) and reload data to continue refining the dataset.", kind='warn')
_s = mo.callout("GO BACK TO STEP 4b) and reload data before continuing.", kind='warn')
_s
return
@@ -437,7 +482,7 @@ def _(mo):
logo_switch = mo.ui.switch(label="Include Chase Logo", value=False)
n_words = mo.ui.slider(start=10, stop=200, step=1, value=40, debounce=True, show_value=True, label="Max number of words in WordCloud")
n_words = mo.ui.slider(start=10, stop=200, step=1, value=100, debounce=True, show_value=True, label="Max number of words in WordCloud")
return buffer, canvas_size, logo_switch, n_words