1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
##
## name: Pause Poor performing Keywords with 0 Conv. and QS <5
## description:
##  Look at the creation Date and make sure the KW is at least 30 days old. Then, if Conversions are 0 and Quality score is <5 pause the KW
## 
## author: Dana Waidhas, Michael S. Huang
## created: 2024-04-02
## 


########### START - Local Mode Config ###########
# Step 1: Uncomment download_preview_input flag and run Preview successfully with the Datasources you want
download_preview_input=False
# Step 2: In MarinOne, go to Scripts -> Preview -> Logs, download 'dataSourceDict' pickle file, and update pickle_path below
# pickle_path = ''
pickle_path = '/Users/mhuang/Downloads/pickle/roc_media_keyword_qs_20240410_no_conv_filter.pkl'
# Step 3: Copy this script into local IDE with Python virtual env loaded with pandas and numpy.
# Step 4: Run locally with below code to init dataSourceDict

# determine if code is running on server or locally
def is_executing_on_server():
    try:
        # Attempt to access a known restricted builtin
        dict_items = dataSourceDict.items()
        return True
    except NameError:
        # NameError: dataSourceDict object is missing (indicating not on server)
        return False

local_dev = False

if is_executing_on_server():
    print("Code is executing on server. Skip init.")
elif len(pickle_path) > 3:
    print("Code is NOT executing on server. Doing init.")
    local_dev = True
    # load dataSourceDict via pickled file
    import pickle
    dataSourceDict = pickle.load(open(pickle_path, 'rb'))

    # print shape and first 5 rows for each entry in dataSourceDict
    for key, value in dataSourceDict.items():
        print(f"Shape of dataSourceDict[{key}]: {value.shape}")
        # print(f"First 5 rows of dataSourceDict[{key}]:\n{value.head(5)}")

    # set outputDf same as inputDf
    inputDf = dataSourceDict["1"]
    outputDf = inputDf.copy()

    # setup timezone
    import datetime
  # Chicago Timezone is GMT-5. Adjust as needed.
    CLIENT_TIMEZONE = datetime.timezone(datetime.timedelta(hours=-5))

    # import pandas
    import pandas as pd
    import numpy as np

    # other imports
    import re
    import urllib

    # import Marin util functions
    from marin_scripts_utils import tableize, select_changed

    # pandas settings
    pd.set_option('display.max_columns', None)  # Display all columns
    pd.set_option('display.max_colwidth', None)  # Display full content of each column

else:
    print("Running locally but no pickle path defined. dataSourceDict not loaded.")
    exit(1)
########### END - Local Mode Setup ###########


MATCH_TYPE = {
  'EXACT': 'exact',
  'PHRASE': 'phrase',
  'BROAD': 'broad',
}

# Setup timezone
CLIENT_TIMEZONE = datetime.timezone(datetime.timedelta(hours=-4))

# primary data source and columns
inputDf = dataSourceDict["1"]
RPT_COL_KEYWORD = 'Keyword'
RPT_COL_DATE = 'Date'
RPT_COL_STATUS = 'Status'
RPT_COL_PUBLISHER = 'Publisher'
RPT_COL_ACCOUNT = 'Account'
RPT_COL_CAMPAIGN = 'Campaign'
RPT_COL_GROUP = 'Group'
RPT_COL_MATCH_TYPE = 'Match Type'
RPT_COL_HIST_QS = 'Hist. QS'
RPT_COL_QUALITY_SCORE = 'Quality Score'
RPT_COL_IMPR = 'Impr.'
RPT_COL_CONV = 'Conv.'
RPT_COL_CREATION_DATE = 'Creation Date'

# output columns and initial values
BULK_COL_ACCOUNT = 'Account'
BULK_COL_CAMPAIGN = 'Campaign'
BULK_COL_GROUP = 'Group'
BULK_COL_KEYWORD = 'Keyword'
BULK_COL_MATCH_TYPE = 'Match Type'
BULK_COL_STATUS = 'Status'
BULK_COL_AUTOMATION_INFO = 'AUTOMATION INFO'
outputDf[BULK_COL_STATUS] = "<<YOUR VALUE>>"
outputDf[BULK_COL_AUTOMATION_INFO] = "<<YOUR VALUE>>"

VAL_STATUS_PAUSED = 'Paused'
VAL_STATUS_ACTIVE = 'Active'
VAL_PUBLISHER_GOOGLE = 'Google'

print("inputDf.shape", inputDf.shape)
print(inputDf.info())

### user code starts here

groupby_cols = [RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN, RPT_COL_GROUP, RPT_COL_KEYWORD, RPT_COL_MATCH_TYPE]

## Data Cleanup
inputDf_reduced = inputDf.drop(columns=[RPT_COL_STATUS]) \
                         .set_index(groupby_cols)

# Ensure RPT_COL_CREATION_DATE is a Pandas DateTime type with the correct format
inputDf_reduced[RPT_COL_CREATION_DATE] = inputDf_reduced[RPT_COL_CREATION_DATE].str.replace('a.m.', 'AM').str.replace('p.m.', 'PM')
inputDf_reduced[RPT_COL_CREATION_DATE] = pd.to_datetime(inputDf_reduced[RPT_COL_CREATION_DATE])
inputDf_reduced[RPT_COL_DATE] = pd.to_datetime(inputDf_reduced[RPT_COL_DATE])

print(f"Report Date Range: {inputDf_reduced[RPT_COL_DATE].min().date()} - {inputDf_reduced[RPT_COL_DATE].max().date()}")

# DEBUG: pick a specific keyword. Format is (kw, matchtype)
# TEST_KWS = [('new townhouses for sale near me','Broad')]
# group_keys = list(inputDf.groupby(groupby_cols).groups.keys())
# test_keys = [key for key in group_keys if any((test_kw[0] in key[3]) & (test_kw[1] in key[4]) for test_kw in TEST_KWS)]
# inputDf_reduced = inputDf_reduced.loc[test_keys]

print("inputDf_reduced.shape", inputDf_reduced.shape)
print(inputDf_reduced.info())
print("inputDf example", inputDf.head())

## Intermediate calculations 
#  - first_traffic_date: first date with traffic in report
#  - effective_creation_date: Creation Date is date created in Marin, so adjust it if there is traffic prior to that

# Group inputDf by Account, Campaign, Group, Keyword, and Match Type
grouped = inputDf_reduced.groupby(groupby_cols)

# Iterate through each group to determine the earliest date with Impressions
# Add this date to a new column called first_traffic_date
def get_first_traffic_date(group):
    group_with_impressions = group[group[RPT_COL_IMPR] > 0]
    if not group_with_impressions.empty:
        return group_with_impressions[RPT_COL_DATE].min()
    else:
        return None

# Apply the function to each group to get first_traffic_date
first_traffic_date_series = grouped.apply(get_first_traffic_date)

# Custom mean function to ignore zeros in Hist QS
def mean_without_zeros(series):
    non_zero_values = series[series != 0]
    if len(non_zero_values) == 0:
        return 0  # or return None if you prefer
    return non_zero_values.mean()

agg_spec = {
    RPT_COL_PUBLISHER: 'last',
    RPT_COL_IMPR: 'sum',
    RPT_COL_CONV: 'sum',
    RPT_COL_HIST_QS: mean_without_zeros, # Hist QS is zero on days without impressions, ignore them when calc mean
    RPT_COL_QUALITY_SCORE: 'last',
    RPT_COL_CREATION_DATE: 'last'
}

# with first_traffic_date calculated, remove Date column by aggregating
df_keywords = grouped.agg(agg_spec)

# Map the first_traffic_date to the original DataFrame using the same groupby columns
df_keywords['first_traffic_date'] = df_keywords.index.map(first_traffic_date_series)

# Compare first_traffic_date and RPT_COL_CREATION_DATE, take the earlier date, and add to new column effective_creation_date
df_keywords['effective_creation_date'] = df_keywords.apply(
    lambda row: min(row['first_traffic_date'], row[RPT_COL_CREATION_DATE]).date() 
    if pd.notnull(row['first_traffic_date']) else row[RPT_COL_CREATION_DATE].date(), 
    axis=1
)


## Find low-performing keywords to pause
#  - at least 30-days old
#  - no Conversions
#  - and one of below:
#   - Hist QS between 0 and 5, non-inclusive
#   - Hist QS is 0, but Quality Score is not zero (to exclude publishers that don't support QS)
#   - both Hist QS and Quality Score are zero, but only for Google

thirty_days_ago = pd.Timestamp.today(CLIENT_TIMEZONE).date() - pd.Timedelta(days=30)
print(f"30 days ago: {thirty_days_ago}")
at_least_30_days_old = df_keywords['effective_creation_date'] < thirty_days_ago
no_conversions = df_keywords[RPT_COL_CONV] == 0
low_hist_qs = (df_keywords[RPT_COL_HIST_QS] > 0) & (df_keywords[RPT_COL_HIST_QS] < 5)
zero_hist_qs = (df_keywords[RPT_COL_HIST_QS] == 0) & (df_keywords[RPT_COL_QUALITY_SCORE] > 0)
zero_hist_qs_and_qs_google = (df_keywords[RPT_COL_HIST_QS] == 0) & (df_keywords[RPT_COL_QUALITY_SCORE] == 0) & (df_keywords[RPT_COL_PUBLISHER] == VAL_PUBLISHER_GOOGLE)

# apply business rules
df_pause = df_keywords.loc[
    at_least_30_days_old &
    no_conversions &
    (low_hist_qs | zero_hist_qs | zero_hist_qs_and_qs_google)
]

## Prepare output

debugDf = df_keywords.reset_index()
output_cols = [RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN, RPT_COL_GROUP, RPT_COL_KEYWORD, RPT_COL_MATCH_TYPE]
outputDf = df_pause.reset_index()[output_cols].copy()

# outputDf[BULK_COL_STATUS] = VAL_STATUS_PAUSED
info_str = f"Paused by Scripts on {pd.Timestamp.today(CLIENT_TIMEZONE).date()} due to low QS"
outputDf[BULK_COL_AUTOMATION_INFO] = info_str

print("outputDf.shape", outputDf.shape)
print("outputDf example", outputDf.head())

## local debug
if local_dev:
    output_filename = 'outputDf.csv'
    outputDf.to_csv(output_filename, index=False)
    print(f"Local Dev: Output written to: {output_filename}")

    debug_filename = 'debugDf.csv'
    debugDf.to_csv(debug_filename, index=False)
    print(f"Local Dev: Debug written to: {debug_filename}")
else:
    print("outputDf", outputDf.head(5))