Skip to content

Instantly share code, notes, and snippets.

View FedericoV's full-sized avatar

Federico Vaggi FedericoV

28 followers · 3 following
View GitHub Profile
@FedericoV
FedericoV / Lasso vs subgradient
Created August 24, 2020 06:04
from sklearn.datasets import make_regression
from sklearn.linear_model import Lasso
import numpy as np
from scipy.optimize import approx_fprime, check_grad, minimize
X, y, w_true = make_regression(n_samples=200, n_features=500, random_state=0, coef=True)
w_init = np.ones_like(w_true)
lam = 10
def l(w, X, y, lam=10):
@FedericoV
FedericoV / gist:bef857ccbe3ee07e99767968bdd40b61
Created May 14, 2020 16:01
def _lambdifygenerated(_Dummy_169, _Dummy_168):
[[beta_0_0, beta_0_1, beta_0_2, beta_0_3, beta_0_4, beta_0_5, beta_0_6, beta_0_7, beta_0_8, beta_0_9, beta_0_10, beta_0_11, beta_0_12, beta_0_13, beta_0_14, beta_0_15, beta_0_16, beta_0_17, beta_0_18, beta_0_19, beta_0_20], [beta_1_0, beta_1_1, beta_1_2, beta_1_3, beta_1_4, beta_1_5, beta_1_6, beta_1_7, beta_1_8, beta_1_9, beta_1_10, beta_1_11, beta_1_12, beta_1_13, beta_1_14, beta_1_15, beta_1_16, beta_1_17, beta_1_18, beta_1_19, beta_1_20], [beta_2_0, beta_2_1, beta_2_2, beta_2_3, beta_2_4, beta_2_5, beta_2_6, beta_2_7, beta_2_8, beta_2_9, beta_2_10, beta_2_11, beta_2_12, beta_2_13, beta_2_14, beta_2_15, beta_2_16, beta_2_17, beta_2_18, beta_2_19, beta_2_20], [beta_3_0, beta_3_1, beta_3_2, beta_3_3, beta_3_4, beta_3_5, beta_3_6, beta_3_7, beta_3_8, beta_3_9, beta_3_10, beta_3_11, beta_3_12, beta_3_13, beta_3_14, beta_3_15, beta_3_16, beta_3_17, beta_3_18, beta_3_19, beta_3_20], [beta_4_0, beta_4_1, beta_4_2, beta_4_3, beta_4_4, beta_4_5, beta_4_6, beta_4_7
@FedericoV
FedericoV / gist:b1a1901a3fba5484edb51bcbe797cfb4
Created February 13, 2020 19:42
feature_hierarchies: Optional[Union[hierarchy_type], List[hierarchy_type]]
if feature_hierarchies:
if not isinstance(feature_hierarchies, list):
graph_hierarchies = [graph_from_hierarchy(feature_hierarchies)]
else:
graph_hierarchies = [graph_from_hierarchy(h) for h in feature_hierarchies]
@FedericoV
FedericoV / Pyre Debugging
Created May 14, 2018 22:15
Debugging Pyre
186590d34d15:FBA_Analytics_Utils vaggi$ pyre --debug check
2018-05-14 15:13:57,354 DEBUG No configuration found at `/Users/vaggi/Code_Libraries/FBA_Analytics_Utils/.pyre_configuration.local`.
2018-05-14 15:13:57,354 DEBUG No configuration found at `.pyre_configuration.local`.
2018-05-14 15:13:57,354 DEBUG Reading configuration `.pyre_configuration`...
2018-05-14 15:13:57,355 DEBUG Found source directories `.`
2018-05-14 15:13:57,356 DEBUG Running `/Users/vaggi/anaconda3/bin/pyre.bin check -debug -sequential -project-root /Users/vaggi/Code_Libraries/FBA_Analytics_Utils -workers 1 -search-path /Users/vaggi/anaconda3/lib/pyre_check/typeshed/stdlib/ .`
2018-05-14 15:13:57,728 ERROR Client exited with error code -5:
@FedericoV
FedericoV / gist:30c1b2f20764bc6092c70a5759c1c33c
Created October 22, 2016 21:17
vol*(0.00577622650467*Ydj1*kb - (Ydj1*kb + kd*vol)*(kd + kr + 0.00577622650467))*(Ydj1*kb + 0.069314718056*vol)**2*(Ydj1*kb + kd*vol)*((YC*kr + 0.069314718056*YC + YP*kd + YP*kr - 0.00577622650467*Ydj1 + 2*Ys*vol)/vol - (-Ydj1*cln3*kb - Ydj1*kb*prot + vol*(YC*kr + 0.069314718056*YC + YP*kd + YP*kr - 0.00577622650467*Ydj1 + Ys*vol))/vol**2)*(Ydj1*cln3*kb**2/(vol**2*(-Ydj1*kb/vol - 0.069314718056)) + cln3*kb/vol)/(Ydj1*cln3*kb**2*(0.00577622650467*Ydj1*kb - (kr + 0.07509094456067)*(Ydj1*kb + 0.069314718056*vol))*(0.00577622650467*Ydj1*kb - (Ydj1*kb + kd*vol)*(kd + kr + 0.00577622650467))*(Ydj1*kb + kd*vol) + Ydj1*kb**2*prot*(0.00577622650467*Ydj1*kb - (kr + 0.07509094456067)*(Ydj1*kb + 0.069314718056*vol))*(0.00577622650467*Ydj1*kb - (Ydj1*kb + kd*vol)*(kd + kr + 0.00577622650467))*(Ydj1*kb + 0.069314718056*vol) - 0.069314718056*cln3*kb*vol*(-0.00577622650467*Ydj1*kb + (kr + 0.069314718056)*(Ydj1*kb + 0.069314718056*vol))*(0.00577622650467*Ydj1*kb - (Ydj1*kb + kd*vol)*(kd + kr + 0.00577622650467))*(Ydj1*kb + kd
@FedericoV
FedericoV / Grid_Search Functionality?
Created June 17, 2016 11:13
for train_idx, test_idx in cv:
X_train, y_train = X[train_idx], y[train_idx]
X_test, y_test = X[test_idx], y[test_idx]
pip.fit(X_train, y_train)
y_pred = pip.predict(X_test)
print (roc_auc_score(y_test, y_pred))
0.70326179109
@FedericoV
FedericoV / pg-pong.py
Created June 3, 2016 13:26 — forked from karpathy/pg-pong.py
Training a Neural Network ATARI Pong agent with Policy Gradients from raw pixels
""" Trains an agent with (stochastic) Policy Gradients on Pong. Uses OpenAI Gym. """
import numpy as np
import cPickle as pickle
import gym
# hyperparameters
H = 200 # number of hidden layer neurons
batch_size = 10 # every how many episodes to do a param update?
learning_rate = 1e-4
gamma = 0.99 # discount factor for reward
@FedericoV
FedericoV / gist:05ad9af84a0ff192187c
Created March 25, 2016 20:17
conv_model = Sequential()
conv_model.add(Embedding(max_features, embedding_size, input_length=maxlen))
conv_model.add(Dropout(0.25))
conv_model.add(Convolution1D(nb_filter=nb_filter,
filter_length=filter_length,
border_mode='valid',
activation='relu',
subsample_length=1))
conv_model.add(MaxPooling1D(pool_length=pool_length))
conv_model.add(LSTM(lstm_output_size))
@FedericoV
FedericoV / gist:6533c311bf4423cd7f38
Created March 24, 2016 13:36
2016-03-24T14:28:11.662+0100 connected to: localhost
2016-03-24T14:28:14.661+0100 yasp_dump.December 290.5 MB
2016-03-24T14:28:17.661+0100 yasp_dump.December 581.2 MB
2016-03-24T14:28:20.661+0100 yasp_dump.December 882.1 MB
2016-03-24T14:28:23.661+0100 yasp_dump.December 1.1 GB
2016-03-24T14:28:26.661+0100 yasp_dump.December 1.3 GB
2016-03-24T14:28:29.661+0100 yasp_dump.December 1.6 GB
2016-03-24T14:28:32.661+0100 yasp_dump.December 1.9 GB
2016-03-24T14:28:35.661+0100 yasp_dump.December 2.1 GB
2016-03-24T14:28:38.661+0100 yasp_dump.December 2.3 GB
@FedericoV
FedericoV / gist:eed2ac539558a43f04bd
Created June 17, 2015 14:36
Old Notebook
{
"worksheets": [
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Julia and its Ecosystem\n",
"-------------------------------------\n",