API tools faq
paste
Advertisement
Guest User

Untitled

a guest
Mar 27th, 2017
60
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 74.22 KB | None | 0 0
raw download clone embed print report
  1. {
  2. "cells": [
  3. {
  4. "cell_type": "code",
  5. "execution_count": 1,
  6. "metadata": {
  7. "collapsed": false
  8. },
  9. "outputs": [
  10. {
  11. "data": {
  12. "text/html": [
  13. "<script>jQuery(function() {if (jQuery(\"body.notebook_app\").length == 0) { jQuery(\".input_area\").toggle(); jQuery(\".prompt\").toggle();}});</script>"
  14. ]
  15. },
  16. "metadata": {},
  17. "output_type": "display_data"
  18. },
  19. {
  20. "data": {
  21. "text/html": [
  22. "<button onclick=\"jQuery('.input_area').toggle(); jQuery('.prompt').toggle();\">Toggle code</button>"
  23. ]
  24. },
  25. "metadata": {},
  26. "output_type": "display_data"
  27. }
  28. ],
  29. "source": [
  30. "import IPython.core.display as di\n",
  31. "\n",
  32. "# This line will hide code by default when the notebook is exported as HTML\n",
  33. "di.display_html('<script>jQuery(function() {if (jQuery(\"body.notebook_app\").length == 0) { jQuery(\".input_area\").toggle(); jQuery(\".prompt\").toggle();}});</script>', raw=True)\n",
  34. "\n",
  35. "# This line will add a button to toggle visibility of code blocks, for use with the HTML export version\n",
  36. "di.display_html('''<button onclick=\"jQuery('.input_area').toggle(); jQuery('.prompt').toggle();\">Toggle code</button>''', raw=True)"
  37. ]
  38. },
  39. {
  40. "cell_type": "code",
  41. "execution_count": 2,
  42. "metadata": {
  43. "collapsed": true
  44. },
  45. "outputs": [],
  46. "source": [
  47. "from IPython.display import display, HTML\n",
  48. "from itertools import combinations, product"
  49. ]
  50. },
  51. {
  52. "cell_type": "code",
  53. "execution_count": 3,
  54. "metadata": {
  55. "collapsed": false
  56. },
  57. "outputs": [],
  58. "source": [
  59. "from jarvis.brain.utility.data_preparation import (get_metrics_single_date, get_metrics_date_range)\n",
  60. "import numpy as np\n",
  61. "import pandas as pd\n",
  62. "from sklearn import tree\n",
  63. "import pydotplus\n",
  64. "global_metrics = ['ga_cpt', 'fb_spend', 'ga_transactions', 'ga_revenue']\n",
  65. "goal_metric = 'ga_cpt'"
  66. ]
  67. },
  68. {
  69. "cell_type": "code",
  70. "execution_count": 5,
  71. "metadata": {
  72. "collapsed": false
  73. },
  74. "outputs": [],
  75. "source": [
  76. "start = '2017-02-21'\n",
  77. "end = '2017-03-21'\n",
  78. "df = get_metrics_date_range(119, 'ad', start, end, global_metrics, custom_tags=True, active=False)"
  79. ]
  80. },
  81. {
  82. "cell_type": "code",
  83. "execution_count": 6,
  84. "metadata": {
  85. "collapsed": true
  86. },
  87. "outputs": [],
  88. "source": [
  89. "df_g = df"
  90. ]
  91. },
  92. {
  93. "cell_type": "code",
  94. "execution_count": 7,
  95. "metadata": {
  96. "collapsed": false
  97. },
  98. "outputs": [],
  99. "source": [
  100. "df= df[df['App Type'] == 'Web']"
  101. ]
  102. },
  103. {
  104. "cell_type": "code",
  105. "execution_count": 8,
  106. "metadata": {
  107. "collapsed": true
  108. },
  109. "outputs": [],
  110. "source": [
  111. "excluded_tags = ['Campaigns', 'ad_id', 'Image', 'Adsets', 'start_date', '_id', 'Ads',\n",
  112. " 'end_date', 'Active Ads Count in Adset', 'Active Adsets Count in Campaign', 'Labels', 'audience type', 'audience category', 'Product Set',\n",
  113. " 'Prominent Text', 'Dominant Color', 'DominantColor', 'ProminentText', 'NumberofFaces', 'Number of Faces', 'Sale', 'name', 'Carousel Count' , 'Ad Format']"
  114. ]
  115. },
  116. {
  117. "cell_type": "code",
  118. "execution_count": 9,
  119. "metadata": {
  120. "collapsed": true
  121. },
  122. "outputs": [],
  123. "source": [
  124. "def exclude_single_dominant_tag(tags, df):\n",
  125. " all_tags = []\n",
  126. " total_spend = df.fb_spend.sum()\n",
  127. " for tag in tags:\n",
  128. " grp = df.groupby(tag).sum()\n",
  129. " grp['prop'] = grp.fb_spend / total_spend\n",
  130. " if pd.np.max(grp['prop']) > 0.95:\n",
  131. " print(tag)\n",
  132. " continue\n",
  133. " all_tags.append(tag)\n",
  134. " return all_tags\n",
  135. "def clean_tags(df, excluded_tags, global_metrics):\n",
  136. " all_tags = list(set(df.columns) - set(global_metrics) - set(excluded_tags))\n",
  137. " all_tags = [tag for tag in all_tags if len(df[df[tag] != 'None'][tag].unique()) > 1]\n",
  138. " all_tags = [tag for tag in all_tags if df[tag].count() > 0.25*len(df)]\n",
  139. " all_tags = exclude_single_dominant_tag(all_tags, df) \n",
  140. " return all_tags"
  141. ]
  142. },
  143. {
  144. "cell_type": "code",
  145. "execution_count": 10,
  146. "metadata": {
  147. "collapsed": false
  148. },
  149. "outputs": [
  150. {
  151. "name": "stdout",
  152. "output_type": "stream",
  153. "text": [
  154. "Behaviors\n",
  155. "User OS\n"
  156. ]
  157. }
  158. ],
  159. "source": [
  160. "all_tags = clean_tags(df, excluded_tags, global_metrics)"
  161. ]
  162. },
  163. {
  164. "cell_type": "markdown",
  165. "metadata": {},
  166. "source": [
  167. "### Tags after cleanup"
  168. ]
  169. },
  170. {
  171. "cell_type": "code",
  172. "execution_count": 11,
  173. "metadata": {
  174. "collapsed": false
  175. },
  176. "outputs": [
  177. {
  178. "data": {
  179. "text/plain": [
  180. "['Audience Types',\n",
  181. " 'Gender',\n",
  182. " 'Device Platforms',\n",
  183. " 'Landing Pages',\n",
  184. " 'Lookalike Types',\n",
  185. " 'Facebook Positions',\n",
  186. " 'Audience Strategy',\n",
  187. " 'Interests',\n",
  188. " 'Custom Audiences',\n",
  189. " 'Ad Type',\n",
  190. " 'Locations',\n",
  191. " 'Optimization Goal',\n",
  192. " 'Campaign Objective',\n",
  193. " 'strategy',\n",
  194. " 'Age Range',\n",
  195. " 'Current Adsets',\n",
  196. " 'Billing Event',\n",
  197. " 'AdsetStrategy',\n",
  198. " 'Publisher Platforms']"
  199. ]
  200. },
  201. "execution_count": 11,
  202. "metadata": {},
  203. "output_type": "execute_result"
  204. }
  205. ],
  206. "source": [
  207. "all_tags"
  208. ]
  209. },
  210. {
  211. "cell_type": "code",
  212. "execution_count": 12,
  213. "metadata": {
  214. "collapsed": true
  215. },
  216. "outputs": [],
  217. "source": [
  218. "df_org = df.copy()"
  219. ]
  220. },
  221. {
  222. "cell_type": "code",
  223. "execution_count": 13,
  224. "metadata": {
  225. "collapsed": false
  226. },
  227. "outputs": [],
  228. "source": [
  229. "df = df_org[df_org['App Type'] == 'Web']"
  230. ]
  231. },
  232. {
  233. "cell_type": "code",
  234. "execution_count": 14,
  235. "metadata": {
  236. "collapsed": false
  237. },
  238. "outputs": [],
  239. "source": [
  240. "overall_cpt = df.fb_spend.sum() / df.ga_transactions.sum()"
  241. ]
  242. },
  243. {
  244. "cell_type": "code",
  245. "execution_count": 15,
  246. "metadata": {
  247. "collapsed": true
  248. },
  249. "outputs": [],
  250. "source": [
  251. "df_global = df.copy()"
  252. ]
  253. },
  254. {
  255. "cell_type": "code",
  256. "execution_count": 16,
  257. "metadata": {
  258. "collapsed": false
  259. },
  260. "outputs": [],
  261. "source": [
  262. "total_spend = df_global.fb_spend.sum()"
  263. ]
  264. },
  265. {
  266. "cell_type": "code",
  267. "execution_count": 17,
  268. "metadata": {
  269. "collapsed": false
  270. },
  271. "outputs": [],
  272. "source": [
  273. "tag_scores = {}\n",
  274. "for tag in all_tags:\n",
  275. " temp = df.groupby([tag]).sum()\n",
  276. " temp['prop'] = temp['fb_spend'] / total_spend\n",
  277. " tag_scores[tag] = temp[temp.index != 'None']['prop'].mean()"
  278. ]
  279. },
  280. {
  281. "cell_type": "markdown",
  282. "metadata": {},
  283. "source": [
  284. "### Scores"
  285. ]
  286. },
  287. {
  288. "cell_type": "code",
  289. "execution_count": 18,
  290. "metadata": {
  291. "collapsed": false
  292. },
  293. "outputs": [
  294. {
  295. "data": {
  296. "text/plain": [
  297. "[('Interests', 0.012374212466183187),\n",
  298. " ('Landing Pages', 0.016936776591743516),\n",
  299. " ('Custom Audiences', 0.025144310123355166),\n",
  300. " ('Lookalike Types', 0.0362420481952802),\n",
  301. " ('Age Range', 0.07142857142857138),\n",
  302. " ('Audience Types', 0.12491133392877618),\n",
  303. " ('strategy', 0.1595410051068123),\n",
  304. " ('Facebook Positions', 0.16368058291676418),\n",
  305. " ('Locations', 0.16666666666666674),\n",
  306. " ('Publisher Platforms', 0.2),\n",
  307. " ('Optimization Goal', 0.24999999999999992),\n",
  308. " ('Campaign Objective', 0.24999999999999997),\n",
  309. " ('Current Adsets', 0.26607416518977656),\n",
  310. " ('AdsetStrategy', 0.3059956515591048),\n",
  311. " ('Audience Strategy', 0.3333333333333332),\n",
  312. " ('Gender', 0.33333333333333326),\n",
  313. " ('Device Platforms', 0.3333333333333334),\n",
  314. " ('Ad Type', 0.33333333333333387),\n",
  315. " ('Billing Event', 0.4999999999999998)]"
  316. ]
  317. },
  318. "execution_count": 18,
  319. "metadata": {},
  320. "output_type": "execute_result"
  321. }
  322. ],
  323. "source": [
  324. "import operator\n",
  325. "sorted(tag_scores.items(), key=operator.itemgetter(1))"
  326. ]
  327. },
  328. {
  329. "cell_type": "code",
  330. "execution_count": 19,
  331. "metadata": {
  332. "collapsed": false
  333. },
  334. "outputs": [],
  335. "source": [
  336. "score_df = pd.DataFrame().from_dict(tag_scores, orient='index')\n",
  337. "score_df = score_df.rename(columns = {0:'score'})"
  338. ]
  339. },
  340. {
  341. "cell_type": "code",
  342. "execution_count": 20,
  343. "metadata": {
  344. "collapsed": true
  345. },
  346. "outputs": [],
  347. "source": [
  348. "score_df = score_df[pd.notnull(score_df['score'])]"
  349. ]
  350. },
  351. {
  352. "cell_type": "code",
  353. "execution_count": 21,
  354. "metadata": {
  355. "collapsed": true
  356. },
  357. "outputs": [],
  358. "source": [
  359. "num_levels = 4"
  360. ]
  361. },
  362. {
  363. "cell_type": "code",
  364. "execution_count": 22,
  365. "metadata": {
  366. "collapsed": false
  367. },
  368. "outputs": [],
  369. "source": [
  370. "from sklearn.cluster import KMeans\n",
  371. "from sklearn import preprocessing\n",
  372. "\n",
  373. "train = score_df[pd.notnull(score_df['score'])].values\n",
  374. "train = preprocessing.scale(train)\n",
  375. "kmeans_model = KMeans(n_clusters=num_levels, random_state=1).fit(train)"
  376. ]
  377. },
  378. {
  379. "cell_type": "code",
  380. "execution_count": 23,
  381. "metadata": {
  382. "collapsed": false
  383. },
  384. "outputs": [],
  385. "source": [
  386. "score_df['level'] = kmeans_model.labels_"
  387. ]
  388. },
  389. {
  390. "cell_type": "code",
  391. "execution_count": 24,
  392. "metadata": {
  393. "collapsed": false
  394. },
  395. "outputs": [],
  396. "source": [
  397. "score_levels = {}\n",
  398. "for level in range(num_levels):\n",
  399. " score_levels[level] = score_df[score_df['level'] == level]['score'].mean()\n",
  400. "\n",
  401. "levels = sorted(score_levels.items(), key=operator.itemgetter(1), reverse=True)"
  402. ]
  403. },
  404. {
  405. "cell_type": "code",
  406. "execution_count": 25,
  407. "metadata": {
  408. "collapsed": false
  409. },
  410. "outputs": [],
  411. "source": [
  412. "global_tag_list = [score_df[score_df['level']==levels[level][0]].index.tolist() for level in range(num_levels)]"
  413. ]
  414. },
  415. {
  416. "cell_type": "markdown",
  417. "metadata": {},
  418. "source": [
  419. "### Tag levels"
  420. ]
  421. },
  422. {
  423. "cell_type": "code",
  424. "execution_count": 26,
  425. "metadata": {
  426. "collapsed": false
  427. },
  428. "outputs": [
  429. {
  430. "name": "stdout",
  431. "output_type": "stream",
  432. "text": [
  433. "Level 1: ['Billing Event']\n",
  434. "Level 2: ['Gender', 'Ad Type', 'Current Adsets', 'Device Platforms', 'Audience Strategy', 'Optimization Goal', 'AdsetStrategy', 'Campaign Objective']\n",
  435. "Level 3: ['Locations', 'strategy', 'Audience Types', 'Publisher Platforms', 'Facebook Positions']\n",
  436. "Level 4: ['Custom Audiences', 'Landing Pages', 'Lookalike Types', 'Interests', 'Age Range']\n"
  437. ]
  438. }
  439. ],
  440. "source": [
  441. "for level in range(num_levels):\n",
  442. " print('Level {0}:'.format(level + 1) , global_tag_list[level])"
  443. ]
  444. },
  445. {
  446. "cell_type": "code",
  447. "execution_count": 27,
  448. "metadata": {
  449. "collapsed": true
  450. },
  451. "outputs": [],
  452. "source": [
  453. "# global_tag_list = ['Campaign Objective']"
  454. ]
  455. },
  456. {
  457. "cell_type": "code",
  458. "execution_count": 28,
  459. "metadata": {
  460. "collapsed": true
  461. },
  462. "outputs": [],
  463. "source": [
  464. "aggregations = {'fb_spend': pd.np.sum,\n",
  465. " 'ga_transactions': pd.np.sum,\n",
  466. " 'ga_revenue': pd.np.sum,\n",
  467. " }"
  468. ]
  469. },
  470. {
  471. "cell_type": "code",
  472. "execution_count": 29,
  473. "metadata": {
  474. "collapsed": false
  475. },
  476. "outputs": [],
  477. "source": [
  478. "overall = df_global['fb_spend'].sum() / df_global['ga_transactions'].sum()"
  479. ]
  480. },
  481. {
  482. "cell_type": "code",
  483. "execution_count": 30,
  484. "metadata": {
  485. "collapsed": false
  486. },
  487. "outputs": [],
  488. "source": [
  489. "df_global['ga_cpt'] = df_global['fb_spend'] / df_global['ga_transactions']\n",
  490. "df_global.replace([pd.np.inf, -pd.np.inf, float('inf')], 0, inplace=True)"
  491. ]
  492. },
  493. {
  494. "cell_type": "code",
  495. "execution_count": 31,
  496. "metadata": {
  497. "collapsed": true
  498. },
  499. "outputs": [],
  500. "source": [
  501. "overall_dev = df_global.ga_cpt.std()"
  502. ]
  503. },
  504. {
  505. "cell_type": "code",
  506. "execution_count": 32,
  507. "metadata": {
  508. "collapsed": false
  509. },
  510. "outputs": [],
  511. "source": [
  512. "P = pd.np.percentile(df_global.fb_spend, [10, 100])"
  513. ]
  514. },
  515. {
  516. "cell_type": "code",
  517. "execution_count": 33,
  518. "metadata": {
  519. "collapsed": false
  520. },
  521. "outputs": [],
  522. "source": [
  523. "df_global_true = df_global.copy()"
  524. ]
  525. },
  526. {
  527. "cell_type": "code",
  528. "execution_count": 34,
  529. "metadata": {
  530. "collapsed": true
  531. },
  532. "outputs": [],
  533. "source": [
  534. "df_global = df_global#[(df_global.fb_spend >= P[0]) & (df_global.fb_spend <= P[1])]"
  535. ]
  536. },
  537. {
  538. "cell_type": "code",
  539. "execution_count": 35,
  540. "metadata": {
  541. "collapsed": true
  542. },
  543. "outputs": [],
  544. "source": [
  545. "df_global = df_global[df_global.ga_cpt > 0]"
  546. ]
  547. },
  548. {
  549. "cell_type": "code",
  550. "execution_count": 96,
  551. "metadata": {
  552. "collapsed": false
  553. },
  554. "outputs": [],
  555. "source": [
  556. "def process_paths(df, tag_list_obj):\n",
  557. " \n",
  558. " \n",
  559. " def recurse_paths(tag_list, tag_list_old, tag_values_old):\n",
  560. " if len(tag_list):\n",
  561. " tag_level = tag_list.pop(0)\n",
  562. "\n",
  563. " tags = []\n",
  564. " if isinstance(tag_level, list):\n",
  565. " tags = tags + tag_level\n",
  566. " else:\n",
  567. " tags = tags + [tag_level]\n",
  568. " \n",
  569. " if len(tag_list_old):\n",
  570. " old = df.groupby(tag_list_old).agg(aggregations)\n",
  571. " old['ga_cpt'] = old['fb_spend'] / old['ga_transactions']\n",
  572. " old_cpt = old.loc[tag_values_old]['ga_cpt']\n",
  573. " old_spend = old.loc[tag_values_old]['fb_spend']\n",
  574. " else:\n",
  575. " old = df\n",
  576. " old_cpt = old.fb_spend.sum() / old.ga_transactions.sum()\n",
  577. " old_spend = old.fb_spend.sum()\n",
  578. "\n",
  579. " premature = True\n",
  580. " for tag in tags:\n",
  581. " new_tag_list = tag_list_old + [tag]\n",
  582. "\n",
  583. " new = df.groupby(new_tag_list).agg(aggregations)\n",
  584. " new['ga_cpt'] = new['fb_spend'] / new['ga_transactions']\n",
  585. " prop = new['fb_spend'] / old_spend\n",
  586. " \n",
  587. " if pd.np.max(prop) > 0.95:\n",
  588. " continue\n",
  589. " \n",
  590. " tag_values = set([val[-1] if isinstance(val, tuple) else val for val in new.index.values.tolist()] )\n",
  591. "\n",
  592. " if (tag_values == ['None']):\n",
  593. " continue\n",
  594. "\n",
  595. " for value in tag_values:\n",
  596. " if value == 'None':\n",
  597. " continue\n",
  598. " new_tag_value = tag_values_old + (value,)\n",
  599. " try:\n",
  600. " new.loc[new_tag_value]\n",
  601. " except:\n",
  602. " continue\n",
  603. " if True:#(new.loc[new_tag_value].fb_spend >= P[0]) & (new.loc[new_tag_value].fb_spend <= P[1]):\n",
  604. " value_df = new.loc[new_tag_value]\n",
  605. " new_cpt = value_df.fb_spend / value_df.ga_transactions\n",
  606. " new_spend = value_df.fb_spend\n",
  607. " \n",
  608. " bad_condition = ((new_cpt - old_cpt) / old_cpt ) > 0.1\n",
  609. " good_condition = ((new_cpt - old_cpt) / old_cpt )< -0.1\n",
  610. "# print(bad_condition, old_cpt, value_df.fb_spend / value_df.ga_transactions, new_tag_value)\n",
  611. " if bad_condition:\n",
  612. " bad[new_tag_value] = {'cpt': value_df.fb_spend / value_df.ga_transactions,\n",
  613. " 'spend_prop': value_df.fb_spend*100 / df_global.fb_spend.sum(),\n",
  614. " 'imp_factor': abs((new_cpt - old_cpt)*(new_spend / old_spend)),\n",
  615. " 'tag_list': new_tag_list}\n",
  616. " \n",
  617. " if good_condition:\n",
  618. " good[new_tag_value] = {'cpt': value_df.fb_spend / value_df.ga_transactions,\n",
  619. " 'spend_prop': value_df.fb_spend*100 / df_global.fb_spend.sum(),\n",
  620. " 'imp_factor': abs((new_cpt - old_cpt)*(new_spend / old_spend)),\n",
  621. " 'tag_list': new_tag_list} \n",
  622. " \n",
  623. " premature = False\n",
  624. " recurse_paths(tag_list[:], new_tag_list[:], new_tag_value[:])\n",
  625. " \n",
  626. " if premature and len(tag_list):\n",
  627. " recurse_paths(tag_list[:], tag_list_old, tag_values_old)\n",
  628. " \n",
  629. " recurse_paths(tag_list_obj, [], ())"
  630. ]
  631. },
  632. {
  633. "cell_type": "code",
  634. "execution_count": 97,
  635. "metadata": {
  636. "collapsed": true
  637. },
  638. "outputs": [],
  639. "source": [
  640. "bad = {}\n",
  641. "good = {}"
  642. ]
  643. },
  644. {
  645. "cell_type": "code",
  646. "execution_count": 98,
  647. "metadata": {
  648. "collapsed": false
  649. },
  650. "outputs": [],
  651. "source": [
  652. "tag_list = global_tag_list\n",
  653. "process_paths(df_global, tag_list[:])\n",
  654. "bad_raw = bad.copy()\n",
  655. "good_raw = good.copy()"
  656. ]
  657. },
  658. {
  659. "cell_type": "code",
  660. "execution_count": 114,
  661. "metadata": {
  662. "collapsed": false
  663. },
  664. "outputs": [],
  665. "source": [
  666. "def get_filtered_paths(paths, rev=True):\n",
  667. " filtered_paths = {key: value for (key, value) in paths.items() if value['spend_prop'] > 1}\n",
  668. "\n",
  669. " processed_paths = {}\n",
  670. " for key, value in filtered_paths.items():\n",
  671. " l = list(key)\n",
  672. " none_list = ['-']*len(tag_list)\n",
  673. " none_list[:len(l)] = l\n",
  674. " if len(none_list) > 1:\n",
  675. " new_key = tuple(none_list)\n",
  676. " else:\n",
  677. " new_key = none_list[0]\n",
  678. "\n",
  679. " if tuple(value['tag_list']) in processed_paths:\n",
  680. " processed_paths[tuple(value['tag_list'])][new_key] = {metric: val for metric, val in value.items() if metric != 'tag_list'}\n",
  681. " else:\n",
  682. " processed_paths[tuple(value['tag_list'])] = {}\n",
  683. " processed_paths[tuple(value['tag_list'])][new_key] = {metric: val for metric, val in value.items() if metric != 'tag_list'}\n",
  684. "\n",
  685. " impact_values = {}\n",
  686. " for key, value in processed_paths.items():\n",
  687. " out_df = pd.DataFrame().from_dict(value, orient='index')\n",
  688. " impact = out_df.imp_factor.abs().max()\n",
  689. " impact_values[key] = impact\n",
  690. "\n",
  691. " impacts = sorted(impact_values.items(), key=operator.itemgetter(1), reverse=rev)\n",
  692. " \n",
  693. " return processed_paths, impacts"
  694. ]
  695. },
  696. {
  697. "cell_type": "markdown",
  698. "metadata": {},
  699. "source": [
  700. "### Bad paths"
  701. ]
  702. },
  703. {
  704. "cell_type": "code",
  705. "execution_count": 115,
  706. "metadata": {
  707. "collapsed": false,
  708. "scrolled": false
  709. },
  710. "outputs": [
  711. {
  712. "data": {
  713. "text/html": [
  714. "<div>\n",
  715. "<table border=\"1\" class=\"dataframe\">\n",
  716. " <thead>\n",
  717. " <tr style=\"text-align: right;\">\n",
  718. " <th></th>\n",
  719. " <th></th>\n",
  720. " <th></th>\n",
  721. " <th></th>\n",
  722. " <th>spend_prop</th>\n",
  723. " <th>imp_factor</th>\n",
  724. " <th>cpt</th>\n",
  725. " </tr>\n",
  726. " <tr>\n",
  727. " <th>Billing Event</th>\n",
  728. " <th>Device Platforms</th>\n",
  729. " <th>Audience Types</th>\n",
  730. " <th>-</th>\n",
  731. " <th></th>\n",
  732. " <th></th>\n",
  733. " <th></th>\n",
  734. " </tr>\n",
  735. " </thead>\n",
  736. " <tbody>\n",
  737. " <tr>\n",
  738. " <th>Impressions</th>\n",
  739. " <th>desktop or mobile</th>\n",
  740. " <th>Interests</th>\n",
  741. " <th>-</th>\n",
  742. " <td>17.832051</td>\n",
  743. " <td>316.352766</td>\n",
  744. " <td>1908.278364</td>\n",
  745. " </tr>\n",
  746. " </tbody>\n",
  747. "</table>\n",
  748. "</div>"
  749. ],
  750. "text/plain": [
  751. " spend_prop imp_factor \\\n",
  752. "Billing Event Device Platforms Audience Types - \n",
  753. "Impressions desktop or mobile Interests - 17.832051 316.352766 \n",
  754. "\n",
  755. " cpt \n",
  756. "Billing Event Device Platforms Audience Types - \n",
  757. "Impressions desktop or mobile Interests - 1908.278364 "
  758. ]
  759. },
  760. "metadata": {},
  761. "output_type": "display_data"
  762. },
  763. {
  764. "data": {
  765. "text/html": [
  766. "<div>\n",
  767. "<table border=\"1\" class=\"dataframe\">\n",
  768. " <thead>\n",
  769. " <tr style=\"text-align: right;\">\n",
  770. " <th></th>\n",
  771. " <th></th>\n",
  772. " <th></th>\n",
  773. " <th></th>\n",
  774. " <th>spend_prop</th>\n",
  775. " <th>imp_factor</th>\n",
  776. " <th>cpt</th>\n",
  777. " </tr>\n",
  778. " <tr>\n",
  779. " <th>Billing Event</th>\n",
  780. " <th>AdsetStrategy</th>\n",
  781. " <th>Audience Types</th>\n",
  782. " <th>-</th>\n",
  783. " <th></th>\n",
  784. " <th></th>\n",
  785. " <th></th>\n",
  786. " </tr>\n",
  787. " </thead>\n",
  788. " <tbody>\n",
  789. " <tr>\n",
  790. " <th>Impressions</th>\n",
  791. " <th>visits</th>\n",
  792. " <th>Interests</th>\n",
  793. " <th>-</th>\n",
  794. " <td>11.344311</td>\n",
  795. " <td>303.444918</td>\n",
  796. " <td>1872.699811</td>\n",
  797. " </tr>\n",
  798. " </tbody>\n",
  799. "</table>\n",
  800. "</div>"
  801. ],
  802. "text/plain": [
  803. " spend_prop imp_factor \\\n",
  804. "Billing Event AdsetStrategy Audience Types - \n",
  805. "Impressions visits Interests - 11.344311 303.444918 \n",
  806. "\n",
  807. " cpt \n",
  808. "Billing Event AdsetStrategy Audience Types - \n",
  809. "Impressions visits Interests - 1872.699811 "
  810. ]
  811. },
  812. "metadata": {},
  813. "output_type": "display_data"
  814. },
  815. {
  816. "data": {
  817. "text/html": [
  818. "<div>\n",
  819. "<table border=\"1\" class=\"dataframe\">\n",
  820. " <thead>\n",
  821. " <tr style=\"text-align: right;\">\n",
  822. " <th></th>\n",
  823. " <th></th>\n",
  824. " <th></th>\n",
  825. " <th></th>\n",
  826. " <th>spend_prop</th>\n",
  827. " <th>imp_factor</th>\n",
  828. " <th>cpt</th>\n",
  829. " </tr>\n",
  830. " <tr>\n",
  831. " <th>Billing Event</th>\n",
  832. " <th>Gender</th>\n",
  833. " <th>Facebook Positions</th>\n",
  834. " <th>Landing Pages</th>\n",
  835. " <th></th>\n",
  836. " <th></th>\n",
  837. " <th></th>\n",
  838. " </tr>\n",
  839. " </thead>\n",
  840. " <tbody>\n",
  841. " <tr>\n",
  842. " <th>Impressions</th>\n",
  843. " <th>Female</th>\n",
  844. " <th>feed</th>\n",
  845. " <th>www.koovs.com/women/brand/masaba-x-koovs</th>\n",
  846. " <td>1.243248</td>\n",
  847. " <td>276.168030</td>\n",
  848. " <td>4066.184375</td>\n",
  849. " </tr>\n",
  850. " <tr>\n",
  851. " <th>Link_clicks</th>\n",
  852. " <th>Female</th>\n",
  853. " <th>feed or right_hand_column</th>\n",
  854. " <th>www.koovs.com/women/dresses, www.koovs.com/women/tops/sortby-latest, www.koovs.com/women/shoes/sortby-latest, www.koovs.com/women/jeans/sortby-latest</th>\n",
  855. " <td>1.954581</td>\n",
  856. " <td>285.613044</td>\n",
  857. " <td>2764.402432</td>\n",
  858. " </tr>\n",
  859. " </tbody>\n",
  860. "</table>\n",
  861. "</div>"
  862. ],
  863. "text/plain": [
  864. " spend_prop \\\n",
  865. "Billing Event Gender Facebook Positions Landing Pages \n",
  866. "Impressions Female feed www.koovs.com/women/brand/masaba-x-koovs 1.243248 \n",
  867. "Link_clicks Female feed or right_hand_column www.koovs.com/women/dresses, www.koovs.com/wome... 1.954581 \n",
  868. "\n",
  869. " imp_factor \\\n",
  870. "Billing Event Gender Facebook Positions Landing Pages \n",
  871. "Impressions Female feed www.koovs.com/women/brand/masaba-x-koovs 276.168030 \n",
  872. "Link_clicks Female feed or right_hand_column www.koovs.com/women/dresses, www.koovs.com/wome... 285.613044 \n",
  873. "\n",
  874. " cpt \n",
  875. "Billing Event Gender Facebook Positions Landing Pages \n",
  876. "Impressions Female feed www.koovs.com/women/brand/masaba-x-koovs 4066.184375 \n",
  877. "Link_clicks Female feed or right_hand_column www.koovs.com/women/dresses, www.koovs.com/wome... 2764.402432 "
  878. ]
  879. },
  880. "metadata": {},
  881. "output_type": "display_data"
  882. },
  883. {
  884. "data": {
  885. "text/html": [
  886. "<div>\n",
  887. "<table border=\"1\" class=\"dataframe\">\n",
  888. " <thead>\n",
  889. " <tr style=\"text-align: right;\">\n",
  890. " <th></th>\n",
  891. " <th></th>\n",
  892. " <th></th>\n",
  893. " <th></th>\n",
  894. " <th>spend_prop</th>\n",
  895. " <th>imp_factor</th>\n",
  896. " <th>cpt</th>\n",
  897. " </tr>\n",
  898. " <tr>\n",
  899. " <th>Billing Event</th>\n",
  900. " <th>Current Adsets</th>\n",
  901. " <th>Publisher Platforms</th>\n",
  902. " <th>-</th>\n",
  903. " <th></th>\n",
  904. " <th></th>\n",
  905. " <th></th>\n",
  906. " </tr>\n",
  907. " </thead>\n",
  908. " <tbody>\n",
  909. " <tr>\n",
  910. " <th>Impressions</th>\n",
  911. " <th>Spend 2.5L</th>\n",
  912. " <th>audience_network or facebook</th>\n",
  913. " <th>-</th>\n",
  914. " <td>18.00953</td>\n",
  915. " <td>278.78765</td>\n",
  916. " <td>1529.927873</td>\n",
  917. " </tr>\n",
  918. " </tbody>\n",
  919. "</table>\n",
  920. "</div>"
  921. ],
  922. "text/plain": [
  923. " spend_prop \\\n",
  924. "Billing Event Current Adsets Publisher Platforms - \n",
  925. "Impressions Spend 2.5L audience_network or facebook - 18.00953 \n",
  926. "\n",
  927. " imp_factor \\\n",
  928. "Billing Event Current Adsets Publisher Platforms - \n",
  929. "Impressions Spend 2.5L audience_network or facebook - 278.78765 \n",
  930. "\n",
  931. " cpt \n",
  932. "Billing Event Current Adsets Publisher Platforms - \n",
  933. "Impressions Spend 2.5L audience_network or facebook - 1529.927873 "
  934. ]
  935. },
  936. "metadata": {},
  937. "output_type": "display_data"
  938. },
  939. {
  940. "data": {
  941. "text/html": [
  942. "<div>\n",
  943. "<table border=\"1\" class=\"dataframe\">\n",
  944. " <thead>\n",
  945. " <tr style=\"text-align: right;\">\n",
  946. " <th></th>\n",
  947. " <th></th>\n",
  948. " <th></th>\n",
  949. " <th></th>\n",
  950. " <th>spend_prop</th>\n",
  951. " <th>imp_factor</th>\n",
  952. " <th>cpt</th>\n",
  953. " </tr>\n",
  954. " <tr>\n",
  955. " <th>Billing Event</th>\n",
  956. " <th>AdsetStrategy</th>\n",
  957. " <th>Publisher Platforms</th>\n",
  958. " <th>-</th>\n",
  959. " <th></th>\n",
  960. " <th></th>\n",
  961. " <th></th>\n",
  962. " </tr>\n",
  963. " </thead>\n",
  964. " <tbody>\n",
  965. " <tr>\n",
  966. " <th>Impressions</th>\n",
  967. " <th>roas</th>\n",
  968. " <th>audience_network or facebook</th>\n",
  969. " <th>-</th>\n",
  970. " <td>18.00953</td>\n",
  971. " <td>278.78765</td>\n",
  972. " <td>1529.927873</td>\n",
  973. " </tr>\n",
  974. " </tbody>\n",
  975. "</table>\n",
  976. "</div>"
  977. ],
  978. "text/plain": [
  979. " spend_prop \\\n",
  980. "Billing Event AdsetStrategy Publisher Platforms - \n",
  981. "Impressions roas audience_network or facebook - 18.00953 \n",
  982. "\n",
  983. " imp_factor \\\n",
  984. "Billing Event AdsetStrategy Publisher Platforms - \n",
  985. "Impressions roas audience_network or facebook - 278.78765 \n",
  986. "\n",
  987. " cpt \n",
  988. "Billing Event AdsetStrategy Publisher Platforms - \n",
  989. "Impressions roas audience_network or facebook - 1529.927873 "
  990. ]
  991. },
  992. "metadata": {},
  993. "output_type": "display_data"
  994. },
  995. {
  996. "data": {
  997. "text/html": [
  998. "<div>\n",
  999. "<table border=\"1\" class=\"dataframe\">\n",
  1000. " <thead>\n",
  1001. " <tr style=\"text-align: right;\">\n",
  1002. " <th></th>\n",
  1003. " <th></th>\n",
  1004. " <th></th>\n",
  1005. " <th></th>\n",
  1006. " <th>spend_prop</th>\n",
  1007. " <th>imp_factor</th>\n",
  1008. " <th>cpt</th>\n",
  1009. " </tr>\n",
  1010. " <tr>\n",
  1011. " <th>Billing Event</th>\n",
  1012. " <th>Gender</th>\n",
  1013. " <th>Facebook Positions</th>\n",
  1014. " <th>Age Range</th>\n",
  1015. " <th></th>\n",
  1016. " <th></th>\n",
  1017. " <th></th>\n",
  1018. " </tr>\n",
  1019. " </thead>\n",
  1020. " <tbody>\n",
  1021. " <tr>\n",
  1022. " <th>Impressions</th>\n",
  1023. " <th>Female</th>\n",
  1024. " <th>feed</th>\n",
  1025. " <th>18-40</th>\n",
  1026. " <td>1.243248</td>\n",
  1027. " <td>276.16803</td>\n",
  1028. " <td>4066.184375</td>\n",
  1029. " </tr>\n",
  1030. " </tbody>\n",
  1031. "</table>\n",
  1032. "</div>"
  1033. ],
  1034. "text/plain": [
  1035. " spend_prop imp_factor \\\n",
  1036. "Billing Event Gender Facebook Positions Age Range \n",
  1037. "Impressions Female feed 18-40 1.243248 276.16803 \n",
  1038. "\n",
  1039. " cpt \n",
  1040. "Billing Event Gender Facebook Positions Age Range \n",
  1041. "Impressions Female feed 18-40 4066.184375 "
  1042. ]
  1043. },
  1044. "metadata": {},
  1045. "output_type": "display_data"
  1046. },
  1047. {
  1048. "data": {
  1049. "text/html": [
  1050. "<div>\n",
  1051. "<table border=\"1\" class=\"dataframe\">\n",
  1052. " <thead>\n",
  1053. " <tr style=\"text-align: right;\">\n",
  1054. " <th></th>\n",
  1055. " <th></th>\n",
  1056. " <th></th>\n",
  1057. " <th></th>\n",
  1058. " <th>spend_prop</th>\n",
  1059. " <th>imp_factor</th>\n",
  1060. " <th>cpt</th>\n",
  1061. " </tr>\n",
  1062. " <tr>\n",
  1063. " <th>Billing Event</th>\n",
  1064. " <th>Current Adsets</th>\n",
  1065. " <th>strategy</th>\n",
  1066. " <th>-</th>\n",
  1067. " <th></th>\n",
  1068. " <th></th>\n",
  1069. " <th></th>\n",
  1070. " </tr>\n",
  1071. " </thead>\n",
  1072. " <tbody>\n",
  1073. " <tr>\n",
  1074. " <th>Impressions</th>\n",
  1075. " <th>Spend 2.5L</th>\n",
  1076. " <th>conversions</th>\n",
  1077. " <th>-</th>\n",
  1078. " <td>16.98156</td>\n",
  1079. " <td>264.417576</td>\n",
  1080. " <td>1532.141534</td>\n",
  1081. " </tr>\n",
  1082. " </tbody>\n",
  1083. "</table>\n",
  1084. "</div>"
  1085. ],
  1086. "text/plain": [
  1087. " spend_prop imp_factor \\\n",
  1088. "Billing Event Current Adsets strategy - \n",
  1089. "Impressions Spend 2.5L conversions - 16.98156 264.417576 \n",
  1090. "\n",
  1091. " cpt \n",
  1092. "Billing Event Current Adsets strategy - \n",
  1093. "Impressions Spend 2.5L conversions - 1532.141534 "
  1094. ]
  1095. },
  1096. "metadata": {},
  1097. "output_type": "display_data"
  1098. },
  1099. {
  1100. "data": {
  1101. "text/html": [
  1102. "<div>\n",
  1103. "<table border=\"1\" class=\"dataframe\">\n",
  1104. " <thead>\n",
  1105. " <tr style=\"text-align: right;\">\n",
  1106. " <th></th>\n",
  1107. " <th></th>\n",
  1108. " <th></th>\n",
  1109. " <th></th>\n",
  1110. " <th>spend_prop</th>\n",
  1111. " <th>imp_factor</th>\n",
  1112. " <th>cpt</th>\n",
  1113. " </tr>\n",
  1114. " <tr>\n",
  1115. " <th>Billing Event</th>\n",
  1116. " <th>AdsetStrategy</th>\n",
  1117. " <th>strategy</th>\n",
  1118. " <th>-</th>\n",
  1119. " <th></th>\n",
  1120. " <th></th>\n",
  1121. " <th></th>\n",
  1122. " </tr>\n",
  1123. " </thead>\n",
  1124. " <tbody>\n",
  1125. " <tr>\n",
  1126. " <th>Impressions</th>\n",
  1127. " <th>roas</th>\n",
  1128. " <th>conversions</th>\n",
  1129. " <th>-</th>\n",
  1130. " <td>16.98156</td>\n",
  1131. " <td>264.417576</td>\n",
  1132. " <td>1532.141534</td>\n",
  1133. " </tr>\n",
  1134. " </tbody>\n",
  1135. "</table>\n",
  1136. "</div>"
  1137. ],
  1138. "text/plain": [
  1139. " spend_prop imp_factor cpt\n",
  1140. "Billing Event AdsetStrategy strategy - \n",
  1141. "Impressions roas conversions - 16.98156 264.417576 1532.141534"
  1142. ]
  1143. },
  1144. "metadata": {},
  1145. "output_type": "display_data"
  1146. },
  1147. {
  1148. "data": {
  1149. "text/html": [
  1150. "<div>\n",
  1151. "<table border=\"1\" class=\"dataframe\">\n",
  1152. " <thead>\n",
  1153. " <tr style=\"text-align: right;\">\n",
  1154. " <th></th>\n",
  1155. " <th></th>\n",
  1156. " <th></th>\n",
  1157. " <th></th>\n",
  1158. " <th>spend_prop</th>\n",
  1159. " <th>imp_factor</th>\n",
  1160. " <th>cpt</th>\n",
  1161. " </tr>\n",
  1162. " <tr>\n",
  1163. " <th>Billing Event</th>\n",
  1164. " <th>Gender</th>\n",
  1165. " <th>Audience Types</th>\n",
  1166. " <th>Landing Pages</th>\n",
  1167. " <th></th>\n",
  1168. " <th></th>\n",
  1169. " <th></th>\n",
  1170. " </tr>\n",
  1171. " </thead>\n",
  1172. " <tbody>\n",
  1173. " <tr>\n",
  1174. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1175. " <th rowspan=\"2\" valign=\"top\">Female</th>\n",
  1176. " <th>Interests</th>\n",
  1177. " <th>www.koovs.com/women/brand/masaba-x-koovs</th>\n",
  1178. " <td>1.649678</td>\n",
  1179. " <td>245.149379</td>\n",
  1180. " <td>4316.369000</td>\n",
  1181. " </tr>\n",
  1182. " <tr>\n",
  1183. " <th>Lookalike</th>\n",
  1184. " <th>www.koovs.com/women/dresses, www.koovs.com/women/tops/sortby-latest, www.koovs.com/women/shoes/sortby-latest, www.koovs.com/women/jeans/sortby-latest</th>\n",
  1185. " <td>1.715253</td>\n",
  1186. " <td>179.234914</td>\n",
  1187. " <td>2087.416512</td>\n",
  1188. " </tr>\n",
  1189. " </tbody>\n",
  1190. "</table>\n",
  1191. "</div>"
  1192. ],
  1193. "text/plain": [
  1194. " spend_prop \\\n",
  1195. "Billing Event Gender Audience Types Landing Pages \n",
  1196. "Impressions Female Interests www.koovs.com/women/brand/masaba-x-koovs 1.649678 \n",
  1197. " Lookalike www.koovs.com/women/dresses, www.koovs.com/wome... 1.715253 \n",
  1198. "\n",
  1199. " imp_factor \\\n",
  1200. "Billing Event Gender Audience Types Landing Pages \n",
  1201. "Impressions Female Interests www.koovs.com/women/brand/masaba-x-koovs 245.149379 \n",
  1202. " Lookalike www.koovs.com/women/dresses, www.koovs.com/wome... 179.234914 \n",
  1203. "\n",
  1204. " cpt \n",
  1205. "Billing Event Gender Audience Types Landing Pages \n",
  1206. "Impressions Female Interests www.koovs.com/women/brand/masaba-x-koovs 4316.369000 \n",
  1207. " Lookalike www.koovs.com/women/dresses, www.koovs.com/wome... 2087.416512 "
  1208. ]
  1209. },
  1210. "metadata": {},
  1211. "output_type": "display_data"
  1212. },
  1213. {
  1214. "data": {
  1215. "text/html": [
  1216. "<div>\n",
  1217. "<table border=\"1\" class=\"dataframe\">\n",
  1218. " <thead>\n",
  1219. " <tr style=\"text-align: right;\">\n",
  1220. " <th></th>\n",
  1221. " <th></th>\n",
  1222. " <th></th>\n",
  1223. " <th></th>\n",
  1224. " <th>spend_prop</th>\n",
  1225. " <th>imp_factor</th>\n",
  1226. " <th>cpt</th>\n",
  1227. " </tr>\n",
  1228. " <tr>\n",
  1229. " <th>Billing Event</th>\n",
  1230. " <th>Gender</th>\n",
  1231. " <th>Audience Types</th>\n",
  1232. " <th>Age Range</th>\n",
  1233. " <th></th>\n",
  1234. " <th></th>\n",
  1235. " <th></th>\n",
  1236. " </tr>\n",
  1237. " </thead>\n",
  1238. " <tbody>\n",
  1239. " <tr>\n",
  1240. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1241. " <th>Female</th>\n",
  1242. " <th>Interests</th>\n",
  1243. " <th>18-40</th>\n",
  1244. " <td>1.649678</td>\n",
  1245. " <td>245.149379</td>\n",
  1246. " <td>4316.369000</td>\n",
  1247. " </tr>\n",
  1248. " <tr>\n",
  1249. " <th>Male</th>\n",
  1250. " <th>Interests</th>\n",
  1251. " <th>18-34</th>\n",
  1252. " <td>1.039295</td>\n",
  1253. " <td>228.524231</td>\n",
  1254. " <td>4532.178333</td>\n",
  1255. " </tr>\n",
  1256. " </tbody>\n",
  1257. "</table>\n",
  1258. "</div>"
  1259. ],
  1260. "text/plain": [
  1261. " spend_prop imp_factor \\\n",
  1262. "Billing Event Gender Audience Types Age Range \n",
  1263. "Impressions Female Interests 18-40 1.649678 245.149379 \n",
  1264. " Male Interests 18-34 1.039295 228.524231 \n",
  1265. "\n",
  1266. " cpt \n",
  1267. "Billing Event Gender Audience Types Age Range \n",
  1268. "Impressions Female Interests 18-40 4316.369000 \n",
  1269. " Male Interests 18-34 4532.178333 "
  1270. ]
  1271. },
  1272. "metadata": {},
  1273. "output_type": "display_data"
  1274. },
  1275. {
  1276. "data": {
  1277. "text/html": [
  1278. "<div>\n",
  1279. "<table border=\"1\" class=\"dataframe\">\n",
  1280. " <thead>\n",
  1281. " <tr style=\"text-align: right;\">\n",
  1282. " <th></th>\n",
  1283. " <th></th>\n",
  1284. " <th></th>\n",
  1285. " <th></th>\n",
  1286. " <th>spend_prop</th>\n",
  1287. " <th>imp_factor</th>\n",
  1288. " <th>cpt</th>\n",
  1289. " </tr>\n",
  1290. " <tr>\n",
  1291. " <th>Billing Event</th>\n",
  1292. " <th>Gender</th>\n",
  1293. " <th>Audience Types</th>\n",
  1294. " <th>-</th>\n",
  1295. " <th></th>\n",
  1296. " <th></th>\n",
  1297. " <th></th>\n",
  1298. " </tr>\n",
  1299. " </thead>\n",
  1300. " <tbody>\n",
  1301. " <tr>\n",
  1302. " <th>Impressions</th>\n",
  1303. " <th>Female</th>\n",
  1304. " <th>Interests</th>\n",
  1305. " <th>-</th>\n",
  1306. " <td>16.389444</td>\n",
  1307. " <td>234.799858</td>\n",
  1308. " <td>1880.82625</td>\n",
  1309. " </tr>\n",
  1310. " </tbody>\n",
  1311. "</table>\n",
  1312. "</div>"
  1313. ],
  1314. "text/plain": [
  1315. " spend_prop imp_factor cpt\n",
  1316. "Billing Event Gender Audience Types - \n",
  1317. "Impressions Female Interests - 16.389444 234.799858 1880.82625"
  1318. ]
  1319. },
  1320. "metadata": {},
  1321. "output_type": "display_data"
  1322. },
  1323. {
  1324. "data": {
  1325. "text/html": [
  1326. "<div>\n",
  1327. "<table border=\"1\" class=\"dataframe\">\n",
  1328. " <thead>\n",
  1329. " <tr style=\"text-align: right;\">\n",
  1330. " <th></th>\n",
  1331. " <th></th>\n",
  1332. " <th></th>\n",
  1333. " <th></th>\n",
  1334. " <th>spend_prop</th>\n",
  1335. " <th>imp_factor</th>\n",
  1336. " <th>cpt</th>\n",
  1337. " </tr>\n",
  1338. " <tr>\n",
  1339. " <th>Billing Event</th>\n",
  1340. " <th>Device Platforms</th>\n",
  1341. " <th>Publisher Platforms</th>\n",
  1342. " <th>Landing Pages</th>\n",
  1343. " <th></th>\n",
  1344. " <th></th>\n",
  1345. " <th></th>\n",
  1346. " </tr>\n",
  1347. " </thead>\n",
  1348. " <tbody>\n",
  1349. " <tr>\n",
  1350. " <th>Link_clicks</th>\n",
  1351. " <th>desktop or mobile</th>\n",
  1352. " <th>audience_network or facebook</th>\n",
  1353. " <th>www.koovs.com/women/dresses, www.koovs.com/women/tops/sortby-latest, www.koovs.com/women/shoes/sortby-latest, www.koovs.com/women/jeans/sortby-latest</th>\n",
  1354. " <td>1.954581</td>\n",
  1355. " <td>230.161377</td>\n",
  1356. " <td>2764.402432</td>\n",
  1357. " </tr>\n",
  1358. " </tbody>\n",
  1359. "</table>\n",
  1360. "</div>"
  1361. ],
  1362. "text/plain": [
  1363. " spend_prop \\\n",
  1364. "Billing Event Device Platforms Publisher Platforms Landing Pages \n",
  1365. "Link_clicks desktop or mobile audience_network or facebook www.koovs.com/women/dresses, www.koovs.com/wome... 1.954581 \n",
  1366. "\n",
  1367. " imp_factor \\\n",
  1368. "Billing Event Device Platforms Publisher Platforms Landing Pages \n",
  1369. "Link_clicks desktop or mobile audience_network or facebook www.koovs.com/women/dresses, www.koovs.com/wome... 230.161377 \n",
  1370. "\n",
  1371. " cpt \n",
  1372. "Billing Event Device Platforms Publisher Platforms Landing Pages \n",
  1373. "Link_clicks desktop or mobile audience_network or facebook www.koovs.com/women/dresses, www.koovs.com/wome... 2764.402432 "
  1374. ]
  1375. },
  1376. "metadata": {},
  1377. "output_type": "display_data"
  1378. },
  1379. {
  1380. "data": {
  1381. "text/html": [
  1382. "<div>\n",
  1383. "<table border=\"1\" class=\"dataframe\">\n",
  1384. " <thead>\n",
  1385. " <tr style=\"text-align: right;\">\n",
  1386. " <th></th>\n",
  1387. " <th></th>\n",
  1388. " <th></th>\n",
  1389. " <th></th>\n",
  1390. " <th>spend_prop</th>\n",
  1391. " <th>imp_factor</th>\n",
  1392. " <th>cpt</th>\n",
  1393. " </tr>\n",
  1394. " <tr>\n",
  1395. " <th>Billing Event</th>\n",
  1396. " <th>Gender</th>\n",
  1397. " <th>Audience Types</th>\n",
  1398. " <th>Interests</th>\n",
  1399. " <th></th>\n",
  1400. " <th></th>\n",
  1401. " <th></th>\n",
  1402. " </tr>\n",
  1403. " </thead>\n",
  1404. " <tbody>\n",
  1405. " <tr>\n",
  1406. " <th>Impressions</th>\n",
  1407. " <th>Male</th>\n",
  1408. " <th>Interests</th>\n",
  1409. " <th>Allen Solly or Amazon.com or Beauty or Coupons or Discounts and allowances or Others</th>\n",
  1410. " <td>1.039295</td>\n",
  1411. " <td>228.524231</td>\n",
  1412. " <td>4532.178333</td>\n",
  1413. " </tr>\n",
  1414. " </tbody>\n",
  1415. "</table>\n",
  1416. "</div>"
  1417. ],
  1418. "text/plain": [
  1419. " spend_prop \\\n",
  1420. "Billing Event Gender Audience Types Interests \n",
  1421. "Impressions Male Interests Allen Solly or Amazon.com or Beauty or Coupons ... 1.039295 \n",
  1422. "\n",
  1423. " imp_factor \\\n",
  1424. "Billing Event Gender Audience Types Interests \n",
  1425. "Impressions Male Interests Allen Solly or Amazon.com or Beauty or Coupons ... 228.524231 \n",
  1426. "\n",
  1427. " cpt \n",
  1428. "Billing Event Gender Audience Types Interests \n",
  1429. "Impressions Male Interests Allen Solly or Amazon.com or Beauty or Coupons ... 4532.178333 "
  1430. ]
  1431. },
  1432. "metadata": {},
  1433. "output_type": "display_data"
  1434. },
  1435. {
  1436. "data": {
  1437. "text/html": [
  1438. "<div>\n",
  1439. "<table border=\"1\" class=\"dataframe\">\n",
  1440. " <thead>\n",
  1441. " <tr style=\"text-align: right;\">\n",
  1442. " <th></th>\n",
  1443. " <th></th>\n",
  1444. " <th></th>\n",
  1445. " <th></th>\n",
  1446. " <th>spend_prop</th>\n",
  1447. " <th>imp_factor</th>\n",
  1448. " <th>cpt</th>\n",
  1449. " </tr>\n",
  1450. " <tr>\n",
  1451. " <th>Billing Event</th>\n",
  1452. " <th>Device Platforms</th>\n",
  1453. " <th>Publisher Platforms</th>\n",
  1454. " <th>-</th>\n",
  1455. " <th></th>\n",
  1456. " <th></th>\n",
  1457. " <th></th>\n",
  1458. " </tr>\n",
  1459. " </thead>\n",
  1460. " <tbody>\n",
  1461. " <tr>\n",
  1462. " <th>Impressions</th>\n",
  1463. " <th>desktop or mobile</th>\n",
  1464. " <th>audience_network or facebook</th>\n",
  1465. " <th>-</th>\n",
  1466. " <td>26.106853</td>\n",
  1467. " <td>225.713802</td>\n",
  1468. " <td>1597.856269</td>\n",
  1469. " </tr>\n",
  1470. " </tbody>\n",
  1471. "</table>\n",
  1472. "</div>"
  1473. ],
  1474. "text/plain": [
  1475. " spend_prop \\\n",
  1476. "Billing Event Device Platforms Publisher Platforms - \n",
  1477. "Impressions desktop or mobile audience_network or facebook - 26.106853 \n",
  1478. "\n",
  1479. " imp_factor \\\n",
  1480. "Billing Event Device Platforms Publisher Platforms - \n",
  1481. "Impressions desktop or mobile audience_network or facebook - 225.713802 \n",
  1482. "\n",
  1483. " cpt \n",
  1484. "Billing Event Device Platforms Publisher Platforms - \n",
  1485. "Impressions desktop or mobile audience_network or facebook - 1597.856269 "
  1486. ]
  1487. },
  1488. "metadata": {},
  1489. "output_type": "display_data"
  1490. },
  1491. {
  1492. "data": {
  1493. "text/html": [
  1494. "<div>\n",
  1495. "<table border=\"1\" class=\"dataframe\">\n",
  1496. " <thead>\n",
  1497. " <tr style=\"text-align: right;\">\n",
  1498. " <th></th>\n",
  1499. " <th></th>\n",
  1500. " <th></th>\n",
  1501. " <th></th>\n",
  1502. " <th>spend_prop</th>\n",
  1503. " <th>imp_factor</th>\n",
  1504. " <th>cpt</th>\n",
  1505. " </tr>\n",
  1506. " <tr>\n",
  1507. " <th>Billing Event</th>\n",
  1508. " <th>Audience Strategy</th>\n",
  1509. " <th>Audience Types</th>\n",
  1510. " <th>-</th>\n",
  1511. " <th></th>\n",
  1512. " <th></th>\n",
  1513. " <th></th>\n",
  1514. " </tr>\n",
  1515. " </thead>\n",
  1516. " <tbody>\n",
  1517. " <tr>\n",
  1518. " <th>Impressions</th>\n",
  1519. " <th>Prospecting</th>\n",
  1520. " <th>Interests</th>\n",
  1521. " <th>-</th>\n",
  1522. " <td>28.146539</td>\n",
  1523. " <td>210.452198</td>\n",
  1524. " <td>1907.906062</td>\n",
  1525. " </tr>\n",
  1526. " </tbody>\n",
  1527. "</table>\n",
  1528. "</div>"
  1529. ],
  1530. "text/plain": [
  1531. " spend_prop imp_factor \\\n",
  1532. "Billing Event Audience Strategy Audience Types - \n",
  1533. "Impressions Prospecting Interests - 28.146539 210.452198 \n",
  1534. "\n",
  1535. " cpt \n",
  1536. "Billing Event Audience Strategy Audience Types - \n",
  1537. "Impressions Prospecting Interests - 1907.906062 "
  1538. ]
  1539. },
  1540. "metadata": {},
  1541. "output_type": "display_data"
  1542. }
  1543. ],
  1544. "source": [
  1545. "bad_new, bad_impacts = get_filtered_paths(bad_raw, True)\n",
  1546. "bad_impacts_list = [value[0] for value in bad_impacts]\n",
  1547. "for key in bad_impacts_list[0:15]:\n",
  1548. " if key in list(bad_new.keys()):\n",
  1549. " value = bad_new[key]\n",
  1550. " b = pd.DataFrame().from_dict(value, orient='index')\n",
  1551. " none_list = ['-']*len(tag_list)\n",
  1552. " l = list(key)\n",
  1553. " none_list[:len(l)] = l\n",
  1554. "\n",
  1555. " b.index = b.index.set_names(none_list)\n",
  1556. " b = b[b['imp_factor'] > 0.7*b.imp_factor.max()]\n",
  1557. " display(b)\n",
  1558. " "
  1559. ]
  1560. },
  1561. {
  1562. "cell_type": "markdown",
  1563. "metadata": {},
  1564. "source": [
  1565. "### Good Paths"
  1566. ]
  1567. },
  1568. {
  1569. "cell_type": "code",
  1570. "execution_count": 116,
  1571. "metadata": {
  1572. "collapsed": false,
  1573. "scrolled": false
  1574. },
  1575. "outputs": [
  1576. {
  1577. "data": {
  1578. "text/html": [
  1579. "<div>\n",
  1580. "<table border=\"1\" class=\"dataframe\">\n",
  1581. " <thead>\n",
  1582. " <tr style=\"text-align: right;\">\n",
  1583. " <th></th>\n",
  1584. " <th></th>\n",
  1585. " <th></th>\n",
  1586. " <th></th>\n",
  1587. " <th>spend_prop</th>\n",
  1588. " <th>imp_factor</th>\n",
  1589. " <th>cpt</th>\n",
  1590. " </tr>\n",
  1591. " <tr>\n",
  1592. " <th>Billing Event</th>\n",
  1593. " <th>Device Platforms</th>\n",
  1594. " <th>Locations</th>\n",
  1595. " <th>Age Range</th>\n",
  1596. " <th></th>\n",
  1597. " <th></th>\n",
  1598. " <th></th>\n",
  1599. " </tr>\n",
  1600. " </thead>\n",
  1601. " <tbody>\n",
  1602. " <tr>\n",
  1603. " <th>Impressions</th>\n",
  1604. " <th>desktop or mobile</th>\n",
  1605. " <th>IN</th>\n",
  1606. " <th>18-65</th>\n",
  1607. " <td>11.585272</td>\n",
  1608. " <td>164.974182</td>\n",
  1609. " <td>838.527331</td>\n",
  1610. " </tr>\n",
  1611. " </tbody>\n",
  1612. "</table>\n",
  1613. "</div>"
  1614. ],
  1615. "text/plain": [
  1616. " spend_prop imp_factor \\\n",
  1617. "Billing Event Device Platforms Locations Age Range \n",
  1618. "Impressions desktop or mobile IN 18-65 11.585272 164.974182 \n",
  1619. "\n",
  1620. " cpt \n",
  1621. "Billing Event Device Platforms Locations Age Range \n",
  1622. "Impressions desktop or mobile IN 18-65 838.527331 "
  1623. ]
  1624. },
  1625. "metadata": {},
  1626. "output_type": "display_data"
  1627. },
  1628. {
  1629. "data": {
  1630. "text/html": [
  1631. "<div>\n",
  1632. "<table border=\"1\" class=\"dataframe\">\n",
  1633. " <thead>\n",
  1634. " <tr style=\"text-align: right;\">\n",
  1635. " <th></th>\n",
  1636. " <th></th>\n",
  1637. " <th></th>\n",
  1638. " <th></th>\n",
  1639. " <th>spend_prop</th>\n",
  1640. " <th>imp_factor</th>\n",
  1641. " <th>cpt</th>\n",
  1642. " </tr>\n",
  1643. " <tr>\n",
  1644. " <th>Billing Event</th>\n",
  1645. " <th>AdsetStrategy</th>\n",
  1646. " <th>Audience Types</th>\n",
  1647. " <th>-</th>\n",
  1648. " <th></th>\n",
  1649. " <th></th>\n",
  1650. " <th></th>\n",
  1651. " </tr>\n",
  1652. " </thead>\n",
  1653. " <tbody>\n",
  1654. " <tr>\n",
  1655. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1656. " <th>roas</th>\n",
  1657. " <th>DPA</th>\n",
  1658. " <th>-</th>\n",
  1659. " <td>6.354514</td>\n",
  1660. " <td>123.662525</td>\n",
  1661. " <td>678.633878</td>\n",
  1662. " </tr>\n",
  1663. " <tr>\n",
  1664. " <th>visits</th>\n",
  1665. " <th>Lookalike</th>\n",
  1666. " <th>-</th>\n",
  1667. " <td>4.040858</td>\n",
  1668. " <td>146.426187</td>\n",
  1669. " <td>903.664145</td>\n",
  1670. " </tr>\n",
  1671. " </tbody>\n",
  1672. "</table>\n",
  1673. "</div>"
  1674. ],
  1675. "text/plain": [
  1676. " spend_prop imp_factor \\\n",
  1677. "Billing Event AdsetStrategy Audience Types - \n",
  1678. "Impressions roas DPA - 6.354514 123.662525 \n",
  1679. " visits Lookalike - 4.040858 146.426187 \n",
  1680. "\n",
  1681. " cpt \n",
  1682. "Billing Event AdsetStrategy Audience Types - \n",
  1683. "Impressions roas DPA - 678.633878 \n",
  1684. " visits Lookalike - 903.664145 "
  1685. ]
  1686. },
  1687. "metadata": {},
  1688. "output_type": "display_data"
  1689. },
  1690. {
  1691. "data": {
  1692. "text/html": [
  1693. "<div>\n",
  1694. "<table border=\"1\" class=\"dataframe\">\n",
  1695. " <thead>\n",
  1696. " <tr style=\"text-align: right;\">\n",
  1697. " <th></th>\n",
  1698. " <th></th>\n",
  1699. " <th></th>\n",
  1700. " <th></th>\n",
  1701. " <th>spend_prop</th>\n",
  1702. " <th>imp_factor</th>\n",
  1703. " <th>cpt</th>\n",
  1704. " </tr>\n",
  1705. " <tr>\n",
  1706. " <th>Billing Event</th>\n",
  1707. " <th>AdsetStrategy</th>\n",
  1708. " <th>Facebook Positions</th>\n",
  1709. " <th>-</th>\n",
  1710. " <th></th>\n",
  1711. " <th></th>\n",
  1712. " <th></th>\n",
  1713. " </tr>\n",
  1714. " </thead>\n",
  1715. " <tbody>\n",
  1716. " <tr>\n",
  1717. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1718. " <th>roas</th>\n",
  1719. " <th>feed or right_hand_column</th>\n",
  1720. " <th>-</th>\n",
  1721. " <td>9.085970</td>\n",
  1722. " <td>116.62279</td>\n",
  1723. " <td>840.048092</td>\n",
  1724. " </tr>\n",
  1725. " <tr>\n",
  1726. " <th>visits</th>\n",
  1727. " <th>feed</th>\n",
  1728. " <th>-</th>\n",
  1729. " <td>4.399702</td>\n",
  1730. " <td>143.51486</td>\n",
  1731. " <td>959.315125</td>\n",
  1732. " </tr>\n",
  1733. " </tbody>\n",
  1734. "</table>\n",
  1735. "</div>"
  1736. ],
  1737. "text/plain": [
  1738. " spend_prop \\\n",
  1739. "Billing Event AdsetStrategy Facebook Positions - \n",
  1740. "Impressions roas feed or right_hand_column - 9.085970 \n",
  1741. " visits feed - 4.399702 \n",
  1742. "\n",
  1743. " imp_factor \\\n",
  1744. "Billing Event AdsetStrategy Facebook Positions - \n",
  1745. "Impressions roas feed or right_hand_column - 116.62279 \n",
  1746. " visits feed - 143.51486 \n",
  1747. "\n",
  1748. " cpt \n",
  1749. "Billing Event AdsetStrategy Facebook Positions - \n",
  1750. "Impressions roas feed or right_hand_column - 840.048092 \n",
  1751. " visits feed - 959.315125 "
  1752. ]
  1753. },
  1754. "metadata": {},
  1755. "output_type": "display_data"
  1756. },
  1757. {
  1758. "data": {
  1759. "text/html": [
  1760. "<div>\n",
  1761. "<table border=\"1\" class=\"dataframe\">\n",
  1762. " <thead>\n",
  1763. " <tr style=\"text-align: right;\">\n",
  1764. " <th></th>\n",
  1765. " <th></th>\n",
  1766. " <th></th>\n",
  1767. " <th></th>\n",
  1768. " <th>spend_prop</th>\n",
  1769. " <th>imp_factor</th>\n",
  1770. " <th>cpt</th>\n",
  1771. " </tr>\n",
  1772. " <tr>\n",
  1773. " <th>Billing Event</th>\n",
  1774. " <th>Gender</th>\n",
  1775. " <th>Facebook Positions</th>\n",
  1776. " <th>Age Range</th>\n",
  1777. " <th></th>\n",
  1778. " <th></th>\n",
  1779. " <th></th>\n",
  1780. " </tr>\n",
  1781. " </thead>\n",
  1782. " <tbody>\n",
  1783. " <tr>\n",
  1784. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1785. " <th rowspan=\"2\" valign=\"top\">Female</th>\n",
  1786. " <th>feed</th>\n",
  1787. " <th>18-45</th>\n",
  1788. " <td>9.794822</td>\n",
  1789. " <td>107.971959</td>\n",
  1790. " <td>1081.352996</td>\n",
  1791. " </tr>\n",
  1792. " <tr>\n",
  1793. " <th>feed or right_hand_column</th>\n",
  1794. " <th>18-65</th>\n",
  1795. " <td>3.595755</td>\n",
  1796. " <td>141.969523</td>\n",
  1797. " <td>1213.969355</td>\n",
  1798. " </tr>\n",
  1799. " </tbody>\n",
  1800. "</table>\n",
  1801. "</div>"
  1802. ],
  1803. "text/plain": [
  1804. " spend_prop \\\n",
  1805. "Billing Event Gender Facebook Positions Age Range \n",
  1806. "Impressions Female feed 18-45 9.794822 \n",
  1807. " feed or right_hand_column 18-65 3.595755 \n",
  1808. "\n",
  1809. " imp_factor \\\n",
  1810. "Billing Event Gender Facebook Positions Age Range \n",
  1811. "Impressions Female feed 18-45 107.971959 \n",
  1812. " feed or right_hand_column 18-65 141.969523 \n",
  1813. "\n",
  1814. " cpt \n",
  1815. "Billing Event Gender Facebook Positions Age Range \n",
  1816. "Impressions Female feed 18-45 1081.352996 \n",
  1817. " feed or right_hand_column 18-65 1213.969355 "
  1818. ]
  1819. },
  1820. "metadata": {},
  1821. "output_type": "display_data"
  1822. },
  1823. {
  1824. "data": {
  1825. "text/html": [
  1826. "<div>\n",
  1827. "<table border=\"1\" class=\"dataframe\">\n",
  1828. " <thead>\n",
  1829. " <tr style=\"text-align: right;\">\n",
  1830. " <th></th>\n",
  1831. " <th></th>\n",
  1832. " <th></th>\n",
  1833. " <th></th>\n",
  1834. " <th>spend_prop</th>\n",
  1835. " <th>imp_factor</th>\n",
  1836. " <th>cpt</th>\n",
  1837. " </tr>\n",
  1838. " <tr>\n",
  1839. " <th>Billing Event</th>\n",
  1840. " <th>Gender</th>\n",
  1841. " <th>Facebook Positions</th>\n",
  1842. " <th>Custom Audiences</th>\n",
  1843. " <th></th>\n",
  1844. " <th></th>\n",
  1845. " <th></th>\n",
  1846. " </tr>\n",
  1847. " </thead>\n",
  1848. " <tbody>\n",
  1849. " <tr>\n",
  1850. " <th rowspan=\"2\" valign=\"top\">Impressions</th>\n",
  1851. " <th rowspan=\"2\" valign=\"top\">Female</th>\n",
  1852. " <th>feed</th>\n",
  1853. " <th>Lookalike (IN, 1%) - loyal_purchasers_data_20160415</th>\n",
  1854. " <td>9.731727</td>\n",
  1855. " <td>110.844955</td>\n",
  1856. " <td>1076.658732</td>\n",
  1857. " </tr>\n",
  1858. " <tr>\n",
  1859. " <th>feed or right_hand_column</th>\n",
  1860. " <th>ECOMMERCE MOBILE & EMAIL IDS FEB 2017</th>\n",
  1861. " <td>3.595755</td>\n",
  1862. " <td>141.969523</td>\n",
  1863. " <td>1213.969355</td>\n",
  1864. " </tr>\n",
  1865. " </tbody>\n",
  1866. "</table>\n",
  1867. "</div>"
  1868. ],
  1869. "text/plain": [
  1870. " spend_prop \\\n",
  1871. "Billing Event Gender Facebook Positions Custom Audiences \n",
  1872. "Impressions Female feed Lookalike (IN, 1%) - loyal_purchasers_data_2016... 9.731727 \n",
  1873. " feed or right_hand_column ECOMMERCE MOBILE & EMAIL IDS FEB 2017 3.595755 \n",
  1874. "\n",
  1875. " imp_factor \\\n",
  1876. "Billing Event Gender Facebook Positions Custom Audiences \n",
  1877. "Impressions Female feed Lookalike (IN, 1%) - loyal_purchasers_data_2016... 110.844955 \n",
  1878. " feed or right_hand_column ECOMMERCE MOBILE & EMAIL IDS FEB 2017 141.969523 \n",
  1879. "\n",
  1880. " cpt \n",
  1881. "Billing Event Gender Facebook Positions Custom Audiences \n",
  1882. "Impressions Female feed Lookalike (IN, 1%) - loyal_purchasers_data_2016... 1076.658732 \n",
  1883. " feed or right_hand_column ECOMMERCE MOBILE & EMAIL IDS FEB 2017 1213.969355 "
  1884. ]
  1885. },
  1886. "metadata": {},
  1887. "output_type": "display_data"
  1888. },
  1889. {
  1890. "data": {
  1891. "text/html": [
  1892. "<div>\n",
  1893. "<table border=\"1\" class=\"dataframe\">\n",
  1894. " <thead>\n",
  1895. " <tr style=\"text-align: right;\">\n",
  1896. " <th></th>\n",
  1897. " <th></th>\n",
  1898. " <th></th>\n",
  1899. " <th></th>\n",
  1900. " <th>spend_prop</th>\n",
  1901. " <th>imp_factor</th>\n",
  1902. " <th>cpt</th>\n",
  1903. " </tr>\n",
  1904. " <tr>\n",
  1905. " <th>Billing Event</th>\n",
  1906. " <th>Device Platforms</th>\n",
  1907. " <th>Facebook Positions</th>\n",
  1908. " <th>Landing Pages</th>\n",
  1909. " <th></th>\n",
  1910. " <th></th>\n",
  1911. " <th></th>\n",
  1912. " </tr>\n",
  1913. " </thead>\n",
  1914. " <tbody>\n",
  1915. " <tr>\n",
  1916. " <th>Impressions</th>\n",
  1917. " <th>desktop or mobile</th>\n",
  1918. " <th>feed or right_hand_column</th>\n",
  1919. " <th>koovs.com</th>\n",
  1920. " <td>6.354514</td>\n",
  1921. " <td>136.237021</td>\n",
  1922. " <td>678.633878</td>\n",
  1923. " </tr>\n",
  1924. " </tbody>\n",
  1925. "</table>\n",
  1926. "</div>"
  1927. ],
  1928. "text/plain": [
  1929. " spend_prop \\\n",
  1930. "Billing Event Device Platforms Facebook Positions Landing Pages \n",
  1931. "Impressions desktop or mobile feed or right_hand_column koovs.com 6.354514 \n",
  1932. "\n",
  1933. " imp_factor \\\n",
  1934. "Billing Event Device Platforms Facebook Positions Landing Pages \n",
  1935. "Impressions desktop or mobile feed or right_hand_column koovs.com 136.237021 \n",
  1936. "\n",
  1937. " cpt \n",
  1938. "Billing Event Device Platforms Facebook Positions Landing Pages \n",
  1939. "Impressions desktop or mobile feed or right_hand_column koovs.com 678.633878 "
  1940. ]
  1941. },
  1942. "metadata": {},
  1943. "output_type": "display_data"
  1944. },
  1945. {
  1946. "data": {
  1947. "text/html": [
  1948. "<div>\n",
  1949. "<table border=\"1\" class=\"dataframe\">\n",
  1950. " <thead>\n",
  1951. " <tr style=\"text-align: right;\">\n",
  1952. " <th></th>\n",
  1953. " <th></th>\n",
  1954. " <th></th>\n",
  1955. " <th></th>\n",
  1956. " <th>spend_prop</th>\n",
  1957. " <th>imp_factor</th>\n",
  1958. " <th>cpt</th>\n",
  1959. " </tr>\n",
  1960. " <tr>\n",
  1961. " <th>Billing Event</th>\n",
  1962. " <th>Ad Type</th>\n",
  1963. " <th>strategy</th>\n",
  1964. " <th>Lookalike Types</th>\n",
  1965. " <th></th>\n",
  1966. " <th></th>\n",
  1967. " <th></th>\n",
  1968. " </tr>\n",
  1969. " </thead>\n",
  1970. " <tbody>\n",
  1971. " <tr>\n",
  1972. " <th>Impressions</th>\n",
  1973. " <th>Carousel</th>\n",
  1974. " <th>conversions</th>\n",
  1975. " <th>1%</th>\n",
  1976. " <td>14.408359</td>\n",
  1977. " <td>126.649392</td>\n",
  1978. " <td>1176.266661</td>\n",
  1979. " </tr>\n",
  1980. " </tbody>\n",
  1981. "</table>\n",
  1982. "</div>"
  1983. ],
  1984. "text/plain": [
  1985. " spend_prop imp_factor \\\n",
  1986. "Billing Event Ad Type strategy Lookalike Types \n",
  1987. "Impressions Carousel conversions 1% 14.408359 126.649392 \n",
  1988. "\n",
  1989. " cpt \n",
  1990. "Billing Event Ad Type strategy Lookalike Types \n",
  1991. "Impressions Carousel conversions 1% 1176.266661 "
  1992. ]
  1993. },
  1994. "metadata": {},
  1995. "output_type": "display_data"
  1996. },
  1997. {
  1998. "data": {
  1999. "text/html": [
  2000. "<div>\n",
  2001. "<table border=\"1\" class=\"dataframe\">\n",
  2002. " <thead>\n",
  2003. " <tr style=\"text-align: right;\">\n",
  2004. " <th></th>\n",
  2005. " <th></th>\n",
  2006. " <th></th>\n",
  2007. " <th></th>\n",
  2008. " <th>spend_prop</th>\n",
  2009. " <th>imp_factor</th>\n",
  2010. " <th>cpt</th>\n",
  2011. " </tr>\n",
  2012. " <tr>\n",
  2013. " <th>Billing Event</th>\n",
  2014. " <th>Audience Strategy</th>\n",
  2015. " <th>strategy</th>\n",
  2016. " <th>Lookalike Types</th>\n",
  2017. " <th></th>\n",
  2018. " <th></th>\n",
  2019. " <th></th>\n",
  2020. " </tr>\n",
  2021. " </thead>\n",
  2022. " <tbody>\n",
  2023. " <tr>\n",
  2024. " <th>Impressions</th>\n",
  2025. " <th>Prospecting</th>\n",
  2026. " <th>conversions</th>\n",
  2027. " <th>1%</th>\n",
  2028. " <td>14.408359</td>\n",
  2029. " <td>126.649392</td>\n",
  2030. " <td>1176.266661</td>\n",
  2031. " </tr>\n",
  2032. " </tbody>\n",
  2033. "</table>\n",
  2034. "</div>"
  2035. ],
  2036. "text/plain": [
  2037. " spend_prop \\\n",
  2038. "Billing Event Audience Strategy strategy Lookalike Types \n",
  2039. "Impressions Prospecting conversions 1% 14.408359 \n",
  2040. "\n",
  2041. " imp_factor \\\n",
  2042. "Billing Event Audience Strategy strategy Lookalike Types \n",
  2043. "Impressions Prospecting conversions 1% 126.649392 \n",
  2044. "\n",
  2045. " cpt \n",
  2046. "Billing Event Audience Strategy strategy Lookalike Types \n",
  2047. "Impressions Prospecting conversions 1% 1176.266661 "
  2048. ]
  2049. },
  2050. "metadata": {},
  2051. "output_type": "display_data"
  2052. },
  2053. {
  2054. "data": {
  2055. "text/html": [
  2056. "<div>\n",
  2057. "<table border=\"1\" class=\"dataframe\">\n",
  2058. " <thead>\n",
  2059. " <tr style=\"text-align: right;\">\n",
  2060. " <th></th>\n",
  2061. " <th></th>\n",
  2062. " <th></th>\n",
  2063. " <th></th>\n",
  2064. " <th>spend_prop</th>\n",
  2065. " <th>imp_factor</th>\n",
  2066. " <th>cpt</th>\n",
  2067. " </tr>\n",
  2068. " <tr>\n",
  2069. " <th>Billing Event</th>\n",
  2070. " <th>Campaign Objective</th>\n",
  2071. " <th>strategy</th>\n",
  2072. " <th>Lookalike Types</th>\n",
  2073. " <th></th>\n",
  2074. " <th></th>\n",
  2075. " <th></th>\n",
  2076. " </tr>\n",
  2077. " </thead>\n",
  2078. " <tbody>\n",
  2079. " <tr>\n",
  2080. " <th>Impressions</th>\n",
  2081. " <th>Conversions</th>\n",
  2082. " <th>conversions</th>\n",
  2083. " <th>1%</th>\n",
  2084. " <td>14.408359</td>\n",
  2085. " <td>126.649392</td>\n",
  2086. " <td>1176.266661</td>\n",
  2087. " </tr>\n",
  2088. " </tbody>\n",
  2089. "</table>\n",
  2090. "</div>"
  2091. ],
  2092. "text/plain": [
  2093. " spend_prop \\\n",
  2094. "Billing Event Campaign Objective strategy Lookalike Types \n",
  2095. "Impressions Conversions conversions 1% 14.408359 \n",
  2096. "\n",
  2097. " imp_factor \\\n",
  2098. "Billing Event Campaign Objective strategy Lookalike Types \n",
  2099. "Impressions Conversions conversions 1% 126.649392 \n",
  2100. "\n",
  2101. " cpt \n",
  2102. "Billing Event Campaign Objective strategy Lookalike Types \n",
  2103. "Impressions Conversions conversions 1% 1176.266661 "
  2104. ]
  2105. },
  2106. "metadata": {},
  2107. "output_type": "display_data"
  2108. },
  2109. {
  2110. "data": {
  2111. "text/html": [
  2112. "<div>\n",
  2113. "<table border=\"1\" class=\"dataframe\">\n",
  2114. " <thead>\n",
  2115. " <tr style=\"text-align: right;\">\n",
  2116. " <th></th>\n",
  2117. " <th></th>\n",
  2118. " <th></th>\n",
  2119. " <th></th>\n",
  2120. " <th>spend_prop</th>\n",
  2121. " <th>imp_factor</th>\n",
  2122. " <th>cpt</th>\n",
  2123. " </tr>\n",
  2124. " <tr>\n",
  2125. " <th>Billing Event</th>\n",
  2126. " <th>Device Platforms</th>\n",
  2127. " <th>Locations</th>\n",
  2128. " <th>Landing Pages</th>\n",
  2129. " <th></th>\n",
  2130. " <th></th>\n",
  2131. " <th></th>\n",
  2132. " </tr>\n",
  2133. " </thead>\n",
  2134. " <tbody>\n",
  2135. " <tr>\n",
  2136. " <th>Impressions</th>\n",
  2137. " <th>desktop or mobile</th>\n",
  2138. " <th>IN</th>\n",
  2139. " <th>koovs.com</th>\n",
  2140. " <td>6.354514</td>\n",
  2141. " <td>125.464367</td>\n",
  2142. " <td>678.633878</td>\n",
  2143. " </tr>\n",
  2144. " </tbody>\n",
  2145. "</table>\n",
  2146. "</div>"
  2147. ],
  2148. "text/plain": [
  2149. " spend_prop \\\n",
  2150. "Billing Event Device Platforms Locations Landing Pages \n",
  2151. "Impressions desktop or mobile IN koovs.com 6.354514 \n",
  2152. "\n",
  2153. " imp_factor \\\n",
  2154. "Billing Event Device Platforms Locations Landing Pages \n",
  2155. "Impressions desktop or mobile IN koovs.com 125.464367 \n",
  2156. "\n",
  2157. " cpt \n",
  2158. "Billing Event Device Platforms Locations Landing Pages \n",
  2159. "Impressions desktop or mobile IN koovs.com 678.633878 "
  2160. ]
  2161. },
  2162. "metadata": {},
  2163. "output_type": "display_data"
  2164. }
  2165. ],
  2166. "source": [
  2167. "good_new, good_impacts = get_filtered_paths(good_raw, True)\n",
  2168. "good_impacts_list = [value[0] for value in good_impacts]\n",
  2169. "for key in good_impacts_list[0:10]:\n",
  2170. " if key in list(good_new.keys()):\n",
  2171. " value = good_new[key]\n",
  2172. " b = pd.DataFrame().from_dict(value, orient='index')\n",
  2173. " none_list = ['-']*len(tag_list)\n",
  2174. " l = list(key)\n",
  2175. " none_list[:len(l)] = l\n",
  2176. "\n",
  2177. " b.index = b.index.set_names(none_list)\n",
  2178. " b = b[b['imp_factor'] > 0.7*b.imp_factor.max()]\n",
  2179. "\n",
  2180. " display(b)"
  2181. ]
  2182. },
  2183. {
  2184. "cell_type": "code",
  2185. "execution_count": 117,
  2186. "metadata": {
  2187. "collapsed": false
  2188. },
  2189. "outputs": [],
  2190. "source": [
  2191. "strat_df = df.groupby(['Billing Event', 'Device Platforms', 'Audience Types']).sum()"
  2192. ]
  2193. },
  2194. {
  2195. "cell_type": "code",
  2196. "execution_count": 118,
  2197. "metadata": {
  2198. "collapsed": false
  2199. },
  2200. "outputs": [],
  2201. "source": [
  2202. "strat_df['prop'] = strat_df.fb_spend / df.fb_spend.sum()\n",
  2203. "strat_df['cpt'] = strat_df.fb_spend / strat_df.ga_transactions"
  2204. ]
  2205. },
  2206. {
  2207. "cell_type": "code",
  2208. "execution_count": null,
  2209. "metadata": {
  2210. "collapsed": false
  2211. },
  2212. "outputs": [],
  2213. "source": [
  2214. "strat_df.iloc[strat_df.index.get_level_values('Audience Types') == 'Interests']"
  2215. ]
  2216. },
  2217. {
  2218. "cell_type": "code",
  2219. "execution_count": null,
  2220. "metadata": {
  2221. "collapsed": true
  2222. },
  2223. "outputs": [],
  2224. "source": []
  2225. }
  2226. ],
  2227. "metadata": {
  2228. "anaconda-cloud": {},
  2229. "kernelspec": {
  2230. "display_name": "Python [jarvis1]",
  2231. "language": "python",
  2232. "name": "Python [jarvis1]"
  2233. },
  2234. "language_info": {
  2235. "codemirror_mode": {
  2236. "name": "ipython",
  2237. "version": 3
  2238. },
  2239. "file_extension": ".py",
  2240. "mimetype": "text/x-python",
  2241. "name": "python",
  2242. "nbconvert_exporter": "python",
  2243. "pygments_lexer": "ipython3",
  2244. "version": "3.5.2"
  2245. }
  2246. },
  2247. "nbformat": 4,
  2248. "nbformat_minor": 0
  2249. }
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!