Michelangiolo commited on
Commit
38b18d6
1 Parent(s): 3a9ba56
Files changed (2) hide show
  1. _test.ipynb +14 -239
  2. app.py +7 -2
_test.ipynb CHANGED
@@ -51,7 +51,7 @@
51
  },
52
  {
53
  "cell_type": "code",
54
- "execution_count": 44,
55
  "metadata": {},
56
  "outputs": [
57
  {
@@ -66,7 +66,7 @@
66
  "name": "stdout",
67
  "output_type": "stream",
68
  "text": [
69
- "Running on local URL: http://127.0.0.1:7884\n",
70
  "\n",
71
  "To create a public link, set `share=True` in `launch()`.\n"
72
  ]
@@ -74,7 +74,7 @@
74
  {
75
  "data": {
76
  "text/html": [
77
- "<div><iframe src=\"http://127.0.0.1:7884/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
78
  ],
79
  "text/plain": [
80
  "<IPython.core.display.HTML object>"
@@ -87,238 +87,9 @@
87
  "data": {
88
  "text/plain": []
89
  },
90
- "execution_count": 44,
91
  "metadata": {},
92
  "output_type": "execute_result"
93
- },
94
- {
95
- "data": {
96
- "text/html": [
97
- "<div>\n",
98
- "<style scoped>\n",
99
- " .dataframe tbody tr th:only-of-type {\n",
100
- " vertical-align: middle;\n",
101
- " }\n",
102
- "\n",
103
- " .dataframe tbody tr th {\n",
104
- " vertical-align: top;\n",
105
- " }\n",
106
- "\n",
107
- " .dataframe thead th {\n",
108
- " text-align: right;\n",
109
- " }\n",
110
- "</style>\n",
111
- "<table border=\"1\" class=\"dataframe\">\n",
112
- " <thead>\n",
113
- " <tr style=\"text-align: right;\">\n",
114
- " <th></th>\n",
115
- " <th>name</th>\n",
116
- " <th>raised</th>\n",
117
- " <th>target</th>\n",
118
- " <th>size</th>\n",
119
- " <th>stage</th>\n",
120
- " <th>country</th>\n",
121
- " <th>source</th>\n",
122
- " <th>description</th>\n",
123
- " <th>tags</th>\n",
124
- " </tr>\n",
125
- " </thead>\n",
126
- " <tbody>\n",
127
- " <tr>\n",
128
- " <th>78931</th>\n",
129
- " <td>Developeration</td>\n",
130
- " <td>Undisclosed</td>\n",
131
- " <td>Undisclosed</td>\n",
132
- " <td>11-500+</td>\n",
133
- " <td>c</td>\n",
134
- " <td>sweden</td>\n",
135
- " <td>https://www.startupblink.com</td>\n",
136
- " <td>Developeration AB was founded 2016 and is a st...</td>\n",
137
- " <td>['healthtech']</td>\n",
138
- " </tr>\n",
139
- " <tr>\n",
140
- " <th>77566</th>\n",
141
- " <td>ComplyAdvantage</td>\n",
142
- " <td>Undisclosed</td>\n",
143
- " <td>Undisclosed</td>\n",
144
- " <td>11-500+</td>\n",
145
- " <td>c</td>\n",
146
- " <td>united-kingdom</td>\n",
147
- " <td>https://www.startupblink.com</td>\n",
148
- " <td>We are a financial crime solutions provider co...</td>\n",
149
- " <td>['fintech']</td>\n",
150
- " </tr>\n",
151
- " <tr>\n",
152
- " <th>78674</th>\n",
153
- " <td>Atlas</td>\n",
154
- " <td>Undisclosed</td>\n",
155
- " <td>Undisclosed</td>\n",
156
- " <td>11-500+</td>\n",
157
- " <td>c</td>\n",
158
- " <td>russia</td>\n",
159
- " <td>https://www.startupblink.com</td>\n",
160
- " <td>Atlas Biomedical Holding is developing a netwo...</td>\n",
161
- " <td>['healthtech']</td>\n",
162
- " </tr>\n",
163
- " <tr>\n",
164
- " <th>81682</th>\n",
165
- " <td>48 Factoring Inc</td>\n",
166
- " <td>Undisclosed</td>\n",
167
- " <td>Undisclosed</td>\n",
168
- " <td>11-500+</td>\n",
169
- " <td>c</td>\n",
170
- " <td>united-states</td>\n",
171
- " <td>https://www.startupblink.com</td>\n",
172
- " <td>48 Factoring Inc. is a financial services comp...</td>\n",
173
- " <td>['fintech']</td>\n",
174
- " </tr>\n",
175
- " <tr>\n",
176
- " <th>78926</th>\n",
177
- " <td>Xinca</td>\n",
178
- " <td>Undisclosed</td>\n",
179
- " <td>Undisclosed</td>\n",
180
- " <td>11-500+</td>\n",
181
- " <td>c</td>\n",
182
- " <td>argentina</td>\n",
183
- " <td>https://www.startupblink.com</td>\n",
184
- " <td>Incorporar residuos en la fabricaci&amp;oacute;n d...</td>\n",
185
- " <td>['energy' 'environment']</td>\n",
186
- " </tr>\n",
187
- " <tr>\n",
188
- " <th>...</th>\n",
189
- " <td>...</td>\n",
190
- " <td>...</td>\n",
191
- " <td>...</td>\n",
192
- " <td>...</td>\n",
193
- " <td>...</td>\n",
194
- " <td>...</td>\n",
195
- " <td>...</td>\n",
196
- " <td>...</td>\n",
197
- " <td>...</td>\n",
198
- " </tr>\n",
199
- " <tr>\n",
200
- " <th>80432</th>\n",
201
- " <td>Glow</td>\n",
202
- " <td>Undisclosed</td>\n",
203
- " <td>Undisclosed</td>\n",
204
- " <td>11-500+</td>\n",
205
- " <td>c</td>\n",
206
- " <td>china</td>\n",
207
- " <td>https://www.startupblink.com</td>\n",
208
- " <td>Glow is an ambitious enterprise that uniquely ...</td>\n",
209
- " <td>['healthtech']</td>\n",
210
- " </tr>\n",
211
- " <tr>\n",
212
- " <th>77716</th>\n",
213
- " <td>Owiwi</td>\n",
214
- " <td>Undisclosed</td>\n",
215
- " <td>Undisclosed</td>\n",
216
- " <td>11-500+</td>\n",
217
- " <td>c</td>\n",
218
- " <td>greece</td>\n",
219
- " <td>https://www.startupblink.com</td>\n",
220
- " <td>Owiwi is a fun and engaging psychometric tool ...</td>\n",
221
- " <td>['software' 'data']</td>\n",
222
- " </tr>\n",
223
- " <tr>\n",
224
- " <th>78561</th>\n",
225
- " <td>Quantib</td>\n",
226
- " <td>Undisclosed</td>\n",
227
- " <td>Undisclosed</td>\n",
228
- " <td>11-500+</td>\n",
229
- " <td>c</td>\n",
230
- " <td>the-netherlands</td>\n",
231
- " <td>https://www.startupblink.com</td>\n",
232
- " <td>MRI scan technology to better diagnose -- and ...</td>\n",
233
- " <td>['healthtech']</td>\n",
234
- " </tr>\n",
235
- " <tr>\n",
236
- " <th>77554</th>\n",
237
- " <td>Earnin</td>\n",
238
- " <td>Undisclosed</td>\n",
239
- " <td>Undisclosed</td>\n",
240
- " <td>11-500+</td>\n",
241
- " <td>c</td>\n",
242
- " <td>united-states</td>\n",
243
- " <td>https://www.startupblink.com</td>\n",
244
- " <td>We're building a platform of community-support...</td>\n",
245
- " <td>['fintech']</td>\n",
246
- " </tr>\n",
247
- " <tr>\n",
248
- " <th>80694</th>\n",
249
- " <td>Vibrent Health</td>\n",
250
- " <td>Undisclosed</td>\n",
251
- " <td>Undisclosed</td>\n",
252
- " <td>11-500+</td>\n",
253
- " <td>c</td>\n",
254
- " <td>united-states</td>\n",
255
- " <td>https://www.startupblink.com</td>\n",
256
- " <td>The future of developing new cures for patient...</td>\n",
257
- " <td>['healthtech']</td>\n",
258
- " </tr>\n",
259
- " </tbody>\n",
260
- "</table>\n",
261
- "<p>94 rows × 9 columns</p>\n",
262
- "</div>"
263
- ],
264
- "text/plain": [
265
- " name raised target size stage \\\n",
266
- "78931 Developeration Undisclosed Undisclosed 11-500+ c \n",
267
- "77566 ComplyAdvantage Undisclosed Undisclosed 11-500+ c \n",
268
- "78674 Atlas Undisclosed Undisclosed 11-500+ c \n",
269
- "81682 48 Factoring Inc Undisclosed Undisclosed 11-500+ c \n",
270
- "78926 Xinca Undisclosed Undisclosed 11-500+ c \n",
271
- "... ... ... ... ... ... \n",
272
- "80432 Glow Undisclosed Undisclosed 11-500+ c \n",
273
- "77716 Owiwi Undisclosed Undisclosed 11-500+ c \n",
274
- "78561 Quantib Undisclosed Undisclosed 11-500+ c \n",
275
- "77554 Earnin Undisclosed Undisclosed 11-500+ c \n",
276
- "80694 Vibrent Health Undisclosed Undisclosed 11-500+ c \n",
277
- "\n",
278
- " country source \\\n",
279
- "78931 sweden https://www.startupblink.com \n",
280
- "77566 united-kingdom https://www.startupblink.com \n",
281
- "78674 russia https://www.startupblink.com \n",
282
- "81682 united-states https://www.startupblink.com \n",
283
- "78926 argentina https://www.startupblink.com \n",
284
- "... ... ... \n",
285
- "80432 china https://www.startupblink.com \n",
286
- "77716 greece https://www.startupblink.com \n",
287
- "78561 the-netherlands https://www.startupblink.com \n",
288
- "77554 united-states https://www.startupblink.com \n",
289
- "80694 united-states https://www.startupblink.com \n",
290
- "\n",
291
- " description \\\n",
292
- "78931 Developeration AB was founded 2016 and is a st... \n",
293
- "77566 We are a financial crime solutions provider co... \n",
294
- "78674 Atlas Biomedical Holding is developing a netwo... \n",
295
- "81682 48 Factoring Inc. is a financial services comp... \n",
296
- "78926 Incorporar residuos en la fabricaci&oacute;n d... \n",
297
- "... ... \n",
298
- "80432 Glow is an ambitious enterprise that uniquely ... \n",
299
- "77716 Owiwi is a fun and engaging psychometric tool ... \n",
300
- "78561 MRI scan technology to better diagnose -- and ... \n",
301
- "77554 We're building a platform of community-support... \n",
302
- "80694 The future of developing new cures for patient... \n",
303
- "\n",
304
- " tags \n",
305
- "78931 ['healthtech'] \n",
306
- "77566 ['fintech'] \n",
307
- "78674 ['healthtech'] \n",
308
- "81682 ['fintech'] \n",
309
- "78926 ['energy' 'environment'] \n",
310
- "... ... \n",
311
- "80432 ['healthtech'] \n",
312
- "77716 ['software' 'data'] \n",
313
- "78561 ['healthtech'] \n",
314
- "77554 ['fintech'] \n",
315
- "80694 ['healthtech'] \n",
316
- "\n",
317
- "[94 rows x 9 columns]"
318
- ]
319
- },
320
- "metadata": {},
321
- "output_type": "display_data"
322
  }
323
  ],
324
  "source": [
@@ -345,18 +116,22 @@
345
  " else:\n",
346
  " return x\n",
347
  " df_knn = search(df, query)\n",
348
- " #we live the sorting for last\n",
349
- " df_knn = df_knn.sort_values('raised', ascending=False)\n",
350
  " df_knn['raised'] = df_knn['raised'].apply(lambda x : raised_zero(x))\n",
351
  "\n",
352
  " df_size = filter_df(df_knn, 'size', '==', size, 1000)\n",
353
  " df_target = filter_df(df_size, 'target', 'contains', target, 20)\n",
354
- " df_stage = filter_df(df_target, 'stage', '==', stage.lower(), 10)\n",
 
 
 
 
 
355
  " \n",
356
- " display(df_stage)\n",
357
  " # df_raised = df_target[(df_target['raised'] >= raised) | (df_target['raised'] == 0)]\n",
358
  "\n",
359
- " return df_stage[0:100]\n",
 
360
  "\n",
361
  "with gr.Blocks(theme=gr.themes.Soft(primary_hue='amber', secondary_hue='gray', neutral_hue='amber')) as demo:\n",
362
  " gr.Markdown(\n",
@@ -366,7 +141,7 @@
366
  " )\n",
367
  " size = gr.Radio(['1-10', '11-50', '51-200', '201-500', '500+', '11-500+'], multiselect=False, value='11-500+', label='size')\n",
368
  " target = gr.Radio(['B2B', 'B2C', 'B2G', 'B2B2C'], multiselect=False, value='B2B', label='target')\n",
369
- " stage = gr.Radio(['pre-seed', 'A', 'B', 'C', 'exit'], multiselect=False, value='C', label='stage')\n",
370
  " # raised = gr.Slider(0, 20, value=5, step_size=1, label=\"Minimum raising (in Millions)\")\n",
371
  " query = gr.Textbox(label='Describe the Startup you are searching for', value='age reversing')\n",
372
  " btn = gr.Button(value=\"Search for a Startup\")\n",
 
51
  },
52
  {
53
  "cell_type": "code",
54
+ "execution_count": 47,
55
  "metadata": {},
56
  "outputs": [
57
  {
 
66
  "name": "stdout",
67
  "output_type": "stream",
68
  "text": [
69
+ "Running on local URL: http://127.0.0.1:7887\n",
70
  "\n",
71
  "To create a public link, set `share=True` in `launch()`.\n"
72
  ]
 
74
  {
75
  "data": {
76
  "text/html": [
77
+ "<div><iframe src=\"http://127.0.0.1:7887/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
78
  ],
79
  "text/plain": [
80
  "<IPython.core.display.HTML object>"
 
87
  "data": {
88
  "text/plain": []
89
  },
90
+ "execution_count": 47,
91
  "metadata": {},
92
  "output_type": "execute_result"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
93
  }
94
  ],
95
  "source": [
 
116
  " else:\n",
117
  " return x\n",
118
  " df_knn = search(df, query)\n",
 
 
119
  " df_knn['raised'] = df_knn['raised'].apply(lambda x : raised_zero(x))\n",
120
  "\n",
121
  " df_size = filter_df(df_knn, 'size', '==', size, 1000)\n",
122
  " df_target = filter_df(df_size, 'target', 'contains', target, 20)\n",
123
+ "\n",
124
+ " if stage != 'ALL':\n",
125
+ " df_stage = filter_df(df_target, 'stage', '==', stage.lower(), 10)\n",
126
+ " else:\n",
127
+ " #we bypass the filter\n",
128
+ " df_stage = df_target\n",
129
  " \n",
130
+ " # display(df_stage)\n",
131
  " # df_raised = df_target[(df_target['raised'] >= raised) | (df_target['raised'] == 0)]\n",
132
  "\n",
133
+ " #we live the sorting for last\n",
134
+ " return df_stage[0:100].sort_values('raised', ascending=False)\n",
135
  "\n",
136
  "with gr.Blocks(theme=gr.themes.Soft(primary_hue='amber', secondary_hue='gray', neutral_hue='amber')) as demo:\n",
137
  " gr.Markdown(\n",
 
141
  " )\n",
142
  " size = gr.Radio(['1-10', '11-50', '51-200', '201-500', '500+', '11-500+'], multiselect=False, value='11-500+', label='size')\n",
143
  " target = gr.Radio(['B2B', 'B2C', 'B2G', 'B2B2C'], multiselect=False, value='B2B', label='target')\n",
144
+ " stage = gr.Radio(['pre-seed', 'A', 'B', 'C', 'exit', 'ALL'], multiselect=False, value='ALL', label='stage')\n",
145
  " # raised = gr.Slider(0, 20, value=5, step_size=1, label=\"Minimum raising (in Millions)\")\n",
146
  " query = gr.Textbox(label='Describe the Startup you are searching for', value='age reversing')\n",
147
  " btn = gr.Button(value=\"Search for a Startup\")\n",
app.py CHANGED
@@ -67,7 +67,12 @@ def greet(size, target, stage, query):
67
 
68
  df_size = filter_df(df_knn, 'size', '==', size, 1000)
69
  df_target = filter_df(df_size, 'target', 'contains', target, 20)
70
- df_stage = filter_df(df_target, 'stage', '==', stage.lower(), 10)
 
 
 
 
 
71
 
72
  # display(df_stage)
73
  # df_raised = df_target[(df_target['raised'] >= raised) | (df_target['raised'] == 0)]
@@ -83,7 +88,7 @@ with gr.Blocks(theme=gr.themes.Soft(primary_hue='amber', secondary_hue='gray', n
83
  )
84
  size = gr.Radio(['1-10', '11-50', '51-200', '201-500', '500+', '11-500+'], multiselect=False, value='11-500+', label='size')
85
  target = gr.Radio(['B2B', 'B2C', 'B2G', 'B2B2C'], multiselect=False, value='B2B', label='target')
86
- stage = gr.Radio(['pre-seed', 'A', 'B', 'C', 'exit'], multiselect=False, value='C', label='stage')
87
  # raised = gr.Slider(0, 20, value=5, step_size=1, label="Minimum raising (in Millions)")
88
  query = gr.Textbox(label='Describe the Startup you are searching for', value='age reversing')
89
  btn = gr.Button(value="Search for a Startup")
 
67
 
68
  df_size = filter_df(df_knn, 'size', '==', size, 1000)
69
  df_target = filter_df(df_size, 'target', 'contains', target, 20)
70
+
71
+ if stage != 'ALL':
72
+ df_stage = filter_df(df_target, 'stage', '==', stage.lower(), 10)
73
+ else:
74
+ #we bypass the filter
75
+ df_stage = df_target
76
 
77
  # display(df_stage)
78
  # df_raised = df_target[(df_target['raised'] >= raised) | (df_target['raised'] == 0)]
 
88
  )
89
  size = gr.Radio(['1-10', '11-50', '51-200', '201-500', '500+', '11-500+'], multiselect=False, value='11-500+', label='size')
90
  target = gr.Radio(['B2B', 'B2C', 'B2G', 'B2B2C'], multiselect=False, value='B2B', label='target')
91
+ stage = gr.Radio(['pre-seed', 'A', 'B', 'C', 'exit', 'ALL'], multiselect=False, value='ALL', label='stage')
92
  # raised = gr.Slider(0, 20, value=5, step_size=1, label="Minimum raising (in Millions)")
93
  query = gr.Textbox(label='Describe the Startup you are searching for', value='age reversing')
94
  btn = gr.Button(value="Search for a Startup")