Skip to content
Snippets Groups Projects
Commit aa2cb37f authored by Robert Lanzafame's avatar Robert Lanzafame
Browse files

wip

parent 5c8a4d5d
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
content/GA_2_8/dev/list_of_radial_object_creator.ipynb 0 → 100644
+ 247
− 0
View file @ aa2cb37f
%% Cell type:markdown id: tags:
somehhow this was used to create the pkl file `List_of_radial_object_all_minutes.pkl` or `List_of_radial_object.pkl`
%% Cell type:code id: tags:
``` python
%load_ext autoreload
%autoreload 2
import numpy as np
import matplotlib.pyplot as plt
import scipy.stats as stats
from tickets import *
from minutes import *
from models import *
```
%% Cell type:code id: tags:
``` python
```
%% Cell type:code id: tags:
``` python
counts_2018 = np.genfromtxt('2018_count_250108_0209.csv', delimiter=',', skip_header=1)
day_min_counts = Minutes.day_min(counts_2018[:, 0])
print(f"Loaded {len(counts_2018)} rows of data")
print(f"The shape is {day_min_counts.shape}")
tickets_2018 = np.genfromtxt('2018_tickets_250108_0209.csv', delimiter=',', skip_header=1)
day_min_tickets = Minutes.day_min(tickets_2018)
print(f"Loaded {len(day_min_tickets)} rows of data")
print(f"The shape is {tickets_2018.shape}")
print()
print(f"The mean day is {np.mean(day_min_tickets[:, 0])}")
print(f"The std day is {np.std(day_min_tickets[:, 0])}")
print(f"The mean min is {np.mean(day_min_tickets[:, 1])}")
print(f"The std min is {np.std(day_min_tickets[:, 1])}")
```
%% Output
Loaded 31856 rows of data
The shape is (31856, 2)
Loaded 160813 rows of data
The shape is (160813,)
The mean day is 29.77885494331926
The std day is 8.978221476042602
The mean min is 863.0213913054292
The std min is 255.18749935067723
%% Cell type:code id: tags:
``` python
all = Tickets()
all.add([[1, 60]])
day_min_all = Minutes.day_min(all.tickets)
parameters = [np.mean(day_min_tickets[:, 0]),
np.std(day_min_tickets[:, 0]),
np.mean(day_min_tickets[:, 1]),
np.std(day_min_tickets[:, 1])]
transform = Minutes.get_transform(parameters)
day_c, min_c = transform(day_min_counts[:,0], day_min_counts[:,1])
radius_c = Minutes.radius(day_c, min_c)
day_a, min_a = transform(day_min_all[:,0], day_min_all[:,1])
radius_a = Minutes.radius(day_a, min_a)
radius_max = np.max(radius_a)
print(f"Maximum radius is: {radius_max:.3f}")
```
%% Output
Maximum radius is: 4.772
%% Cell type:code id: tags:
``` python
def get_band_distribution(minute,
radius_c, counts_2018, radius_a, transform,
std_increment=0.125):
day_c, min_c = Minutes.get_day_min(minute[0])
day_t, min_t = transform(day_c, min_c)
radius = Minutes.radius(day_t, min_t)
if isinstance(std_increment, list):
assert len(std_increment) == 2, "std_increment list must have two values"
radius_min = radius - std_increment[0]
radius_max = radius + std_increment[1]
elif isinstance(std_increment, float):
radius_min = radius - std_increment
radius_max = radius + std_increment
else:
raise ValueError("std_increment must be a list of length 2 or a float")
if radius_min < 0:
d_radius = radius_max - radius_min
radius_min = 0
radius_max = d_radius
if radius_max > max(radius_a):
d_radius = radius_max - radius_min
radius_min = radius_min - d_radius
radius_max = max(radius_a)
d = evaluate_ticket_dist_i(radius_min, radius_max,
radius_c, counts_2018[:, 1],
radius_a)
return d
```
%% Cell type:code id: tags:
``` python
# minute = Minutes.get_minutes([['April'], 25, 13, 10])
minute = Minutes.get_minutes([['April'], 29, 13, 30])
print(f"Minute: {minute[0]}")
d = get_band_distribution(minute,
radius_c, counts_2018, radius_a, transform,
std_increment=[.2, .2])
d.hist();
```
%% Output
Minute: 41130
%% Cell type:code id: tags:
``` python
minutes = np.arange(0,60*1440).tolist()
dist_all_minutes = []
for minute in minutes:
dist_all_minutes.append(
get_band_distribution(
[minute], radius_c, counts_2018,
radius_a, transform, std_increment=0.125
)
)
```
%% Cell type:code id: tags:
``` python
with open('pickles/List_of_radial_object_all_minutes.pkl', 'wb') as f:
pickle.dump(dist_all_minutes, f)
print("RadialDist list saved to 'List_of_radial_object_all_minutes.pkl'")
```
%% Output
---------------------------------------------------------------------------
MemoryError Traceback (most recent call last)
Cell In[94], line 2
1 with open('pickles/List_of_radial_object_all_minutes.pkl', 'wb') as f:
----> 2 pickle.dump(dist_all_minutes, f)
4 print("RadialDist list saved to 'List_of_radial_object_all_minutes.pkl'")
MemoryError:
content/GA_2_8/dev/mcs.ipynb
+ 1
− 1
View file @ aa2cb37f
%% Cell type:markdown id: tags:
# MCS
_see markdown file for half-baked theory_
%% Cell type:markdown id: tags:
Constants:
- $V_n$: coefficient of variation of the expected value of the winning ticket
- $N_{s,min}$: minimum number of samples to estimate the expected value of the number of winning tickets
Procedure is:
1. Determine the set of tickets (size $N_t$) to purchase, $T$, defined by minutes $m$
2. Compute total cost $3\cdot N_t$
For each ticket $i$ in $T$ associated with minute $m_i$:
1. Compute probability of being the winning ticket, $p_i=P[M=m_i=m_w]$
2. Sample $N_w(m_i)$; stop when expected value of samples satisfies the criteria defined above (sample index $j=1,\ldots,N_{s,i}$)
3. Compute the profit from each realization $j$ for minute $i$ as $W\cdot N_{w, j}\cdot p_i$
3. Compute the profit from each realization $j$ for minute $i$ as $W \cdot p_i / N_{w, j}$
The samples of $B$ are the collection of all profits computed for the samples of $i$
%% Cell type:code id: tags:
``` python
%load_ext autoreload
%autoreload 2
import numpy as np
import matplotlib.pyplot as plt
import scipy.stats as stats
from tickets import *
from minutes import *
from models import *
```
%% Cell type:code id: tags:
``` python
```
%% Cell type:code id: tags:
``` python
counts_2018 = np.genfromtxt('2018_count_250108_0209.csv', delimiter=',', skip_header=1)
day_min_counts = Minutes.day_min(counts_2018[:, 0])
print(f"Loaded {len(counts_2018)} rows of data")
print(f"The shape is {day_min_counts.shape}")
tickets_2018 = np.genfromtxt('2018_tickets_250108_0209.csv', delimiter=',', skip_header=1)
day_min_tickets = Minutes.day_min(tickets_2018)
print(f"Loaded {len(day_min_tickets)} rows of data")
print(f"The shape is {tickets_2018.shape}")
print()
print(f"The mean day is {np.mean(day_min_tickets[:, 0])}")
print(f"The std day is {np.std(day_min_tickets[:, 0])}")
print(f"The mean min is {np.mean(day_min_tickets[:, 1])}")
print(f"The std min is {np.std(day_min_tickets[:, 1])}")
```
%% Output
Loaded 31856 rows of data
The shape is (31856, 2)
Loaded 160813 rows of data
The shape is (160813,)
The mean day is 29.77885494331926
The std day is 8.978221476042602
The mean min is 863.0213913054292
The std min is 255.18749935067723
%% Cell type:code id: tags:
``` python
all = Tickets()
all.add([[1, 60]])
day_min_all = Minutes.day_min(all.tickets)
parameters = [np.mean(day_min_tickets[:, 0]),
np.std(day_min_tickets[:, 0]),
np.mean(day_min_tickets[:, 1]),
np.std(day_min_tickets[:, 1])]
transform = Minutes.get_transform(parameters)
day_c, min_c = transform(day_min_counts[:,0], day_min_counts[:,1])
radius_c = Minutes.radius(day_c, min_c)
day_a, min_a = transform(day_min_all[:,0], day_min_all[:,1])
radius_a = Minutes.radius(day_a, min_a)
radius_max = np.max(radius_a)
print(f"Maximum radius is: {radius_max:.3f}")
```
%% Output
Maximum radius is: 4.772
%% Cell type:code id: tags:
``` python
def get_band_distribution(minute,
radius_c, counts_2018, radius_a, transform,
std_increment=0.125):
day_c, min_c = Minutes.get_day_min(minute[0])
day_t, min_t = transform(day_c, min_c)
radius = Minutes.radius(day_t, min_t)
if isinstance(std_increment, list):
assert len(std_increment) == 2, "std_increment list must have two values"
radius_min = radius - std_increment[0]
radius_max = radius + std_increment[1]
elif isinstance(std_increment, float):
radius_min = radius - std_increment
radius_max = radius + std_increment
else:
raise ValueError("std_increment must be a list of length 2 or a float")
if radius_min < 0:
d_radius = radius_max - radius_min
radius_min = 0
radius_max = d_radius
if radius_max > max(radius_a):
d_radius = radius_max - radius_min
radius_min = radius_min - d_radius
radius_max = max(radius_a)
d = evaluate_ticket_dist_i(radius_min, radius_max,
radius_c, counts_2018[:, 1],
radius_a)
return d
```
%% Cell type:code id: tags:
``` python
# minute = Minutes.get_minutes([['April'], 25, 13, 10])
minute = Minutes.get_minutes([['April'], 29, 13, 30])
print(f"Minute: {minute[0]}")
d = get_band_distribution(minute,
radius_c, counts_2018, radius_a, transform,
std_increment=[.2, .2])
d.hist();
```
%% Output
Minute: 41130
%% Cell type:code id: tags:
``` python
minutes = np.arange(0,60*1440).tolist()
dist_all_minutes = []
for minute in minutes:
dist_all_minutes.append(
get_band_distribution(
[minute], radius_c, counts_2018,
radius_a, transform, std_increment=0.125
)
)
```
%% Cell type:code id: tags:
``` python
with open('pickles/List_of_radial_object_all_minutes.pkl', 'wb') as f:
pickle.dump(dist_all_minutes, f)
print("RadialDist list saved to 'List_of_radial_object_all_minutes.pkl'")
```
%% Output
---------------------------------------------------------------------------
MemoryError Traceback (most recent call last)
Cell In[94], line 2
1 with open('pickles/List_of_radial_object_all_minutes.pkl', 'wb') as f:
----> 2 pickle.dump(dist_all_minutes, f)
4 print("RadialDist list saved to 'List_of_radial_object_all_minutes.pkl'")
MemoryError:
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment