Skip to content

Commit 605b195

Browse files
MarcoGorelliMarcoGorelli
committed
run black formatter on BDA3
1 parent e27eb30 commit 605b195

File tree

5 files changed

+417
-363
lines changed

5 files changed

+417
-363
lines changed

‎BDA3/chap_02.ipynb

Lines changed: 63 additions & 62 deletions
Original file line numberDiff line numberDiff line change
@@ -35,7 +35,7 @@
3535
"metadata": {},
3636
"outputs": [],
3737
"source": [
38-
"az.style.use('arviz-darkgrid')\n",
38+
"az.style.use(\"arviz-darkgrid\")\n",
3939
"\n",
4040
"%config Inline.figure_formats = ['retina']\n",
4141
"%load_ext watermark"
@@ -58,8 +58,8 @@
5858
"outputs": [],
5959
"source": [
6060
"with pm.Model() as model_1:\n",
61-
" theta = pm.Uniform('theta', lower=0, upper=1)\n",
62-
" obs = pm.Binomial('observed', n=births, p=theta, observed=fem_births)"
61+
" theta = pm.Uniform(\"theta\", lower=0, upper=1)\n",
62+
" obs = pm.Binomial(\"observed\", n=births, p=theta, observed=fem_births)"
6363
]
6464
},
6565
{
@@ -240,7 +240,8 @@
240240
}
241241
],
242242
"source": [
243-
"az.plot_posterior(trace_1); # same as pm.plot_posterior()"
243+
"az.plot_posterior(trace_1)\n",
244+
"# same as pm.plot_posterior()"
244245
]
245246
},
246247
{
@@ -268,10 +269,10 @@
268269
"x = np.linspace(0, 1, 1000)\n",
269270
"y = beta.pdf(x, 438, 544)\n",
270271
"\n",
271-
"mean_t = df['mean'].values[0]\n",
272-
"sd_t = df['sd'].values[0]\n",
273-
"alpha_t = (mean_t**2 * (1 - mean_t)) / (sd_t**2) - mean_t\n",
274-
"beta_t = (1 - mean_t) * (mean_t * (1 - mean_t) / sd_t**2 - 1)\n",
272+
"mean_t = df[\"mean\"].values[0]\n",
273+
"sd_t = df[\"sd\"].values[0]\n",
274+
"alpha_t = (mean_t ** 2 * (1 - mean_t)) / (sd_t ** 2) - mean_t\n",
275+
"beta_t = (1 - mean_t) * (mean_t * (1 - mean_t) / sd_t ** 2 - 1)\n",
275276
"y_pred = beta.pdf(x, alpha_t, beta_t)"
276277
]
277278
},
@@ -293,11 +294,11 @@
293294
],
294295
"source": [
295296
"plt.figure(figsize=(10, 5))\n",
296-
"plt.plot(x, y, label='True', linewidth=5)\n",
297-
"plt.plot(x, y_pred, 'o', label='Predicted', linewidth=4, alpha=0.6)\n",
297+
"plt.plot(x, y, label=\"True\", linewidth=5)\n",
298+
"plt.plot(x, y_pred, \"o\", label=\"Predicted\", linewidth=4, alpha=0.6)\n",
298299
"plt.legend()\n",
299-
"plt.title('The posterior distribution')\n",
300-
"plt.xlabel(r'$\\theta$', fontsize=14);"
300+
"plt.title(\"The posterior distribution\")\n",
301+
"plt.xlabel(r\"$\\theta$\", fontsize=14);"
301302
]
302303
},
303304
{
@@ -314,10 +315,10 @@
314315
"outputs": [],
315316
"source": [
316317
"with pm.Model() as model_2:\n",
317-
" theta = pm.Uniform('theta', lower=0, upper=1)\n",
318-
" trans = pm.Deterministic('trans', pm.logit(theta))\n",
319-
" phi = pm.Deterministic('phi', (1 - theta) / theta)\n",
320-
" obs = pm.Binomial('observed', n=births, p=theta, observed=fem_births)"
318+
" theta = pm.Uniform(\"theta\", lower=0, upper=1)\n",
319+
" trans = pm.Deterministic(\"trans\", pm.logit(theta))\n",
320+
" phi = pm.Deterministic(\"phi\", (1 - theta) / theta)\n",
321+
" obs = pm.Binomial(\"observed\", n=births, p=theta, observed=fem_births)"
321322
]
322323
},
323324
{
@@ -363,10 +364,10 @@
363364
"outputs": [],
364365
"source": [
365366
"with pm.Model() as model_2_bad:\n",
366-
" theta = pm.Uniform('theta', lower=0, upper=1)\n",
367-
" trans = pm.Deterministic('trans', pm.logit(theta))\n",
368-
" phi = pm.Deterministic('phi', (1 - theta) / theta)\n",
369-
" obs = pm.Binomial('observed', n=births, p=theta, observed=-2)"
367+
" theta = pm.Uniform(\"theta\", lower=0, upper=1)\n",
368+
" trans = pm.Deterministic(\"trans\", pm.logit(theta))\n",
369+
" phi = pm.Deterministic(\"phi\", (1 - theta) / theta)\n",
370+
" obs = pm.Binomial(\"observed\", n=births, p=theta, observed=-2)"
370371
]
371372
},
372373
{
@@ -601,7 +602,7 @@
601602
],
602603
"source": [
603604
"fig, axes = plt.subplots(ncols=2, nrows=1, figsize=(11, 4))\n",
604-
"az.plot_posterior(trace_2, var_names=['trans', 'phi'], ax=axes);"
605+
"az.plot_posterior(trace_2, var_names=[\"trans\", \"phi\"], ax=axes);"
605606
]
606607
},
607608
{
@@ -625,10 +626,10 @@
625626
}
626627
],
627628
"source": [
628-
"lldd = expit(df2.loc['trans','hpd_3%'])\n",
629-
"llii = expit(df2.loc['trans','hpd_97%'])\n",
629+
"lldd = expit(df2.loc[\"trans\", \"hpd_3%\"])\n",
630+
"llii = expit(df2.loc[\"trans\", \"hpd_97%\"])\n",
630631
"\n",
631-
"print(f'The interval is [{lldd:.3f}, {llii:.3f}]')"
632+
"print(f\"The interval is [{lldd:.3f}, {llii:.3f}]\")"
632633
]
633634
},
634635
{
@@ -650,18 +651,18 @@
650651
"\n",
651652
"\n",
652653
"def triangular(central_num, width):\n",
653-
" \n",
654+
"\n",
654655
" left_num = central_num - width\n",
655656
" right_num = central_num + width\n",
656-
" theta = pm.Triangular('theta', lower=left_num, upper=right_num, c=central_num)\n",
657-
" \n",
658-
"# Comment these lines to see some changes\n",
657+
" theta = pm.Triangular(\"theta\", lower=left_num, upper=right_num, c=central_num)\n",
658+
"\n",
659+
" # Comment these lines to see some changes\n",
659660
" if tt.lt(left_num, theta):\n",
660-
" theta = pm.Uniform('theta1', lower=0, upper=left_num)\n",
661+
" theta = pm.Uniform(\"theta1\", lower=0, upper=left_num)\n",
661662
" if tt.gt(right_num, theta):\n",
662-
" theta = pm.Uniform('theta2', lower=right_num, upper=1)\n",
663-
" \n",
664-
" return theta "
663+
" theta = pm.Uniform(\"theta2\", lower=right_num, upper=1)\n",
664+
"\n",
665+
" return theta"
665666
]
666667
},
667668
{
@@ -682,7 +683,7 @@
682683
"\n",
683684
"with pm.Model() as model_3:\n",
684685
" theta = triangular(central_num, width)\n",
685-
" obs = pm.Binomial('observed', n=births, p=theta, observed=fem_births)"
686+
" obs = pm.Binomial(\"observed\", n=births, p=theta, observed=fem_births)"
686687
]
687688
},
688689
{
@@ -759,7 +760,7 @@
759760
}
760761
],
761762
"source": [
762-
"az.plot_trace(trace_3, var_names=['theta']);"
763+
"az.plot_trace(trace_3, var_names=[\"theta\"]);"
763764
]
764765
},
765766
{
@@ -834,7 +835,7 @@
834835
}
835836
],
836837
"source": [
837-
"az.summary(trace_3, var_names='theta', round_to=4)"
838+
"az.summary(trace_3, var_names=\"theta\", round_to=4)"
838839
]
839840
},
840841
{
@@ -861,7 +862,7 @@
861862
}
862863
],
863864
"source": [
864-
"az.plot_posterior(trace_3, var_names='theta');"
865+
"az.plot_posterior(trace_3, var_names=\"theta\");"
865866
]
866867
},
867868
{
@@ -878,8 +879,8 @@
878879
"outputs": [],
879880
"source": [
880881
"with pm.Model() as poisson_model:\n",
881-
" theta = pm.Gamma('theta', alpha=3, beta=5)\n",
882-
" post = pm.Poisson('post', mu=2 * theta, observed=3)"
882+
" theta = pm.Gamma(\"theta\", alpha=3, beta=5)\n",
883+
" post = pm.Poisson(\"post\", mu=2 * theta, observed=3)"
883884
]
884885
},
885886
{
@@ -1160,13 +1161,13 @@
11601161
"outputs": [],
11611162
"source": [
11621163
"x = np.linspace(0, 3, 1000)\n",
1163-
"y = gamma.pdf(x, 6, scale=1/7)\n",
1164+
"y = gamma.pdf(x, 6, scale=1 / 7)\n",
11641165
"\n",
1165-
"mean_t = df4['mean'].values[0]\n",
1166-
"sd_t = df4['sd'].values[0]\n",
1167-
"alpha_t = mean_t**2 / sd_t**2\n",
1168-
"beta_t = mean_t / sd_t**2\n",
1169-
"y_pred = gamma.pdf(x, alpha_t, scale=1/beta_t)"
1166+
"mean_t = df4[\"mean\"].values[0]\n",
1167+
"sd_t = df4[\"sd\"].values[0]\n",
1168+
"alpha_t = mean_t ** 2 / sd_t ** 2\n",
1169+
"beta_t = mean_t / sd_t ** 2\n",
1170+
"y_pred = gamma.pdf(x, alpha_t, scale=1 / beta_t)"
11701171
]
11711172
},
11721173
{
@@ -1187,11 +1188,11 @@
11871188
],
11881189
"source": [
11891190
"plt.figure(figsize=(10, 5))\n",
1190-
"plt.plot(x, y, 'k', label='True', linewidth=7)\n",
1191-
"plt.plot(x, y_pred, 'C1', label='Predicted', linewidth=3, alpha=0.9)\n",
1191+
"plt.plot(x, y, \"k\", label=\"True\", linewidth=7)\n",
1192+
"plt.plot(x, y_pred, \"C1\", label=\"Predicted\", linewidth=3, alpha=0.9)\n",
11921193
"plt.legend()\n",
1193-
"plt.title('The posterior distribution')\n",
1194-
"plt.xlabel(r'$\\theta$', fontsize=14);"
1194+
"plt.title(\"The posterior distribution\")\n",
1195+
"plt.xlabel(r\"$\\theta$\", fontsize=14);"
11951196
]
11961197
},
11971198
{
@@ -1208,8 +1209,8 @@
12081209
"outputs": [],
12091210
"source": [
12101211
"with pm.Model() as poisson_model_2:\n",
1211-
" theta = pm.Gamma('theta', alpha=3, beta=5)\n",
1212-
" post = pm.Poisson('post', mu=20 * theta, observed=30)"
1212+
" theta = pm.Gamma(\"theta\", alpha=3, beta=5)\n",
1213+
" post = pm.Poisson(\"post\", mu=20 * theta, observed=30)"
12131214
]
12141215
},
12151216
{
@@ -1399,13 +1400,13 @@
13991400
"outputs": [],
14001401
"source": [
14011402
"x = np.linspace(0, 3, 1000)\n",
1402-
"y = gamma.pdf(x, 33, scale=1/25) # How you write alpha and beta\n",
1403+
"y = gamma.pdf(x, 33, scale=1 / 25) # How you write alpha and beta\n",
14031404
"\n",
1404-
"mean_t = df5['mean'].values[0]\n",
1405-
"sd_t = df5['sd'].values[0]\n",
1406-
"alpha_t = mean_t**2 / sd_t**2\n",
1407-
"beta_t = mean_t / sd_t**2\n",
1408-
"y_pred = gamma.pdf(x, alpha_t, scale=1/beta_t)"
1405+
"mean_t = df5[\"mean\"].values[0]\n",
1406+
"sd_t = df5[\"sd\"].values[0]\n",
1407+
"alpha_t = mean_t ** 2 / sd_t ** 2\n",
1408+
"beta_t = mean_t / sd_t ** 2\n",
1409+
"y_pred = gamma.pdf(x, alpha_t, scale=1 / beta_t)"
14091410
]
14101411
},
14111412
{
@@ -1426,11 +1427,11 @@
14261427
],
14271428
"source": [
14281429
"plt.figure(figsize=(10, 5))\n",
1429-
"plt.plot(x, y, 'k', label='True', linewidth=5)\n",
1430-
"plt.plot(x, y_pred, 'oC1', label='Predicted', alpha=0.15)\n",
1430+
"plt.plot(x, y, \"k\", label=\"True\", linewidth=5)\n",
1431+
"plt.plot(x, y_pred, \"oC1\", label=\"Predicted\", alpha=0.15)\n",
14311432
"plt.legend()\n",
1432-
"plt.title('The posterior distribution')\n",
1433-
"plt.xlabel(r'$\\theta$', fontsize=14);"
1433+
"plt.title(\"The posterior distribution\")\n",
1434+
"plt.xlabel(r\"$\\theta$\", fontsize=14);"
14341435
]
14351436
},
14361437
{
@@ -1447,8 +1448,8 @@
14471448
}
14481449
],
14491450
"source": [
1450-
"val = np.mean(trace_poisson_2['theta'] >= 1)\n",
1451-
"print(f'The posterior probability that theta exceeds 1.0 is {val:.2f}.')"
1451+
"val = np.mean(trace_poisson_2[\"theta\"] >= 1)\n",
1452+
"print(f\"The posterior probability that theta exceeds 1.0 is {val:.2f}.\")"
14521453
]
14531454
},
14541455
{

0 commit comments

Comments
 (0)