القائمة
Walk backward through a time series and multiplicatively rescale segments when jumps exceed a fraction of the running mean to correct data collection anomalies
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Memory-efficient Keras generator that streams sharded CSV files in chunks, renders strokes to images on-the-fly, and yields batches for training on datasets too large for memory
Stack 1D convolutions for local feature extraction before bidirectional LSTMs to classify variable-length stroke sequences into hundreds of doodle categories
Partition a large dataset into N balanced shards using integer key modulo arithmetic for reproducible, class-interleaved splits across CSV files
Normalize raw stroke coordinates to 0-255 range, resample at uniform arc-length spacing, then apply Ramer-Douglas-Peucker simplification
Render stroke sequences to grayscale images with temporal intensity encoding where earlier strokes are brighter and later strokes fade to encode drawing order
Score each face in a video as anomalous by computing L2 distance from the embedding centroid of all faces, then convert to probability via logistic function
| name | timeseries-retroactive-outlier-rescaling |
| description | Walk backward through a time series and multiplicatively rescale segments when jumps exceed a fraction of the running mean to correct data collection anomalies |
Administrative time series (business counts, registrations, economic indicators) sometimes contain abrupt level shifts caused by data collection changes rather than real trends. Walking backward from the most recent value and rescaling earlier segments when the jump exceeds a threshold (e.g., 20% of running mean) corrects these artifacts while preserving genuine trends.
import numpy as np
def fix_outliers(series, threshold_frac=0.20):
values = series.values.copy().astype(float)
n = len(values)
for i in range(n - 2, -1, -1):
running_mean = np.mean(values[i + 1 : min(i + 4, n)])
if running_mean == 0:
continue
jump = abs(values[i] - values[i + 1])
if jump > threshold_frac * running_mean:
scale = values[i + 1] / values[i] if values[i] != 0 else 1.0
values[: i + 1] *= scale
return values
df["value_clean"] = df.groupby("entity_id")["value"].transform(fix_outliers)
|value[i] - value[i+1]| exceeds threshold_frac * running_mean, compute scale factorvalue[i+1] / value[i]