diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 218e1bed254..b8402e84ba3 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -5,11 +5,13 @@ stages:
 
 test_ci:
   stage: test
-  image: ubuntu:18.04
+  image: hub.meri.dev/test-docker/test:latest
 #  only:
 #    - setup-ci
   before_script:
-    - apt update && apt install -y openssh-client wget libarchive-dev libcurl4-openssl-dev git python3.7 python3-pip
+    - apt update && apt install -y libarchive-dev #libcurl4-openssl-dev 
+    - apt install -y zlib1g-dev libicu-dev libcurl3 libcurl-openssl1.0-dev
+    - apt install -y build-essential cmake libssl-dev pkg-config cmake
     - wget http://131.123.42.38/lmcrs/beta/srcML-Ubuntu18.04.deb
     - dpkg -i srcML-Ubuntu18.04.deb
     - mkdir -p ~/.ssh
@@ -22,11 +24,9 @@ test_ci:
     - export LC_ALL=C.UTF-8
     - export LANG=C.UTF-8
   script:
-    - apt-get update
     - git config --global user.email "merico@meri.co"
     - git config --global user.name "merico"
     - pip3 install pipenv
-
     - echo -e "machine gitlab.com\nlogin ${GITLAB_USER}\npassword ${GITLAB_PASSWD}" > ~/.netrc
     - git clone https://gitlab.com/persper/code-analytics.git && cd code-analytics 
     #&& git checkout ${CI_COMMIT_REF_NAME}
diff --git a/Pipfile b/Pipfile
index 19b18641463..dbdd78f493b 100644
--- a/Pipfile
+++ b/Pipfile
@@ -25,6 +25,7 @@ aenum = "*"
 pytest-cov = "*"
 gitpython = "*"
 sphinx = "*"
+python-louvain = "*"
 
 [dev-packages]
 
diff --git a/README.md b/README.md
index 5cebce8dea4..77d7eed63db 100644
--- a/README.md
+++ b/README.md
@@ -15,13 +15,13 @@ The following procedure is tested on Ubuntu 16.04 LTS.
 Download and install Python 3.6+: <https://www.python.org/downloads/>.
 
 Also, create a symbolic link from `python3` to `python` since some scripts reply on it.
-```
+```sh
 sudo ln -s /usr/bin/python3 /usr/bin/python
 ```
 
 2. Install python dependencies (we recommend to use pipenv)
 
-```bash
+```sh
 pipenv install
 ```
 
@@ -29,7 +29,7 @@ pipenv install
 
 In order to uset the `--indent-heuristic` option of `git diff`, we require git version >= 2.11. Use the following commands to upgrade:
 
-```bash
+```sh
 sudo add-apt-repository ppa:git-core/ppa -y
 sudo apt-get update
 sudo apt-get install git -y
@@ -40,12 +40,12 @@ git --version
 
 Add the following line to your `~/.bashrc` file.
 
-```
+```sh
 export PYTHONPATH=$PYTHONPATH:/path/to/dir
 ```
 
 To update your path for the remainder of the session.
-```
+```sh
 source ~/.bashrc
 ```
 
@@ -55,14 +55,21 @@ Please download from [here](https://www.srcml.org/#download) and follow the [ins
 
 srcML also needs `libarchive-dev` and `libcurl4-openssl-dev`. Install them with the following commands:
 
-```bash
+```sh
 sudo apt install libarchive-dev
 sudo apt install libcurl4-openssl-dev
 ```
 
 6. Check setup correctness
 
-```bash
+As the test process will create Git repositories, set up your global Git user name and email before testing:
+```sh
+git config --global user.email "you@example.com"
+git config --global user.name "Your Name"
+```
+
+Run the test process:
+```sh
 pipenv run pytest test/test_analytics
 ```
 
diff --git a/notebooks/demo.ipynb b/notebooks/demo.ipynb
new file mode 100644
index 00000000000..48ddf3760fb
--- /dev/null
+++ b/notebooks/demo.ipynb
@@ -0,0 +1,72 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# import deps\n",
+    "import os\n",
+    "from persper.analytics.c import CGraphServer\n",
+    "from persper.analytics.analyzer2 import Analyzer\n",
+    "from persper.analytics.graph_server import C_FILENAME_REGEXES\n",
+    "from persper.util.path import root_path"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# configure your project\n",
+    "repo_path = os.path.join(root_path, 'repos/<your_repo_name>')\n",
+    "\n",
+    "# configure alpha for devrank\n",
+    "alpha = 0.5"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# start analysis and show commit devrank values\n",
+    "az = Analyzer(repo_path, CGraphServer(C_FILENAME_REGEXES))\n",
+    "await az.analyze()\n",
+    "ccgraph = az.graph\n",
+    "ccgraph.commit_devranks(alpha)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "code-analytics-8iDyuztf",
+   "language": "python",
+   "name": "code-analytics-8idyuztf"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.6.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/persper/analytics/analyzer2.py b/persper/analytics/analyzer2.py
index ecfc398e850..189af7f2899 100644
--- a/persper/analytics/analyzer2.py
+++ b/persper/analytics/analyzer2.py
@@ -2,6 +2,7 @@
 import collections.abc
 import logging
 import re
+import sys
 import time
 from abc import ABC
 from typing import List, Optional, Set, Union, Dict
@@ -22,7 +23,8 @@ def __init__(self, repositoryRoot: str, graphServer: GraphServer,
                  firstParentOnly: bool = False,
                  commit_classifier: Optional[CommitClassifier] = None,
                  skip_rewind_diff: bool = False,
-                 monolithic_commit_lines_threshold: int = 5000):
+                 monolithic_commit_lines_threshold: int = 5000,
+                 monolithic_file_bytes_threshold: int = 200000):
         # skip_rewind_diff will skip diff, but rewind commit start/end will still be notified to the GraphServer.
         self._repositoryRoot = repositoryRoot
         self._graphServer = graphServer
@@ -37,6 +39,8 @@ def __init__(self, repositoryRoot: str, graphServer: GraphServer,
         self._clf_results: Dict[str, List[float]] = {}
         self._skip_rewind_diff = skip_rewind_diff
         self._monolithic_commit_lines_threshold = monolithic_commit_lines_threshold
+        self._monolithic_file_bytes_threshold = monolithic_file_bytes_threshold
+        self._call_commit_graph = None
 
     def __getstate__(self):
         state = self.__dict__.copy()
@@ -107,8 +111,22 @@ def firstParentOnly(self, value: bool):
 
     @property
     def graph(self):
-        return self._graphServer.get_graph()
-
+        # When starting the analysis,set self._call_commit_graph to None, we can ensure that the graph is the latest call commit graph version.
+        if self._call_commit_graph is None:
+            # retry 10 times when get graph from graph server
+            for i in range(10):
+                try:
+                    ccg = self._graphServer.get_graph()
+                    if ccg is not None:
+                        break
+                except Exception:
+                    logging.info('get graph failed:{}'.format(i))
+                    time.sleep(1)
+                    continue
+            else:
+                raise Exception('get graph is failed')
+            self._call_commit_graph = ccg
+        return self._call_commit_graph
     @property
     def visitedCommits(self) -> Set[str]:
         """
@@ -137,7 +155,18 @@ def compute_project_complexity(self, r_n: int, r_e: int):
         """
         return self.graph.eval_project_complexity(r_n, r_e)
 
+    def compute_modularity(self):
+        """Compute modularity score based on function graph.
+
+        Returns
+        -------
+            modularity : float
+                The modularity score of this graph.
+        """
+        return self.graph.compute_modularity()
+
     async def analyze(self, maxAnalyzedCommits=None, suppressStdOutLogs=False):
+        self._call_commit_graph = None
         commitSpec = self._terminalCommit
         if self._originCommit:
             commitSpec = self._originCommit.hexsha + ".." + self._terminalCommit.hexsha
@@ -173,7 +202,7 @@ def printCommitStatus(level, status: str):
                 else:
                     expectedParentCommit = None
                     message = None
-                    if not commit.parents:
+                    if len(commit.parents) == 0:
                         message = "Going forward (initial commit)."
                         expectedParentCommit = None
                     else:
@@ -232,18 +261,22 @@ async def _analyzeCommit(self, commit: Union[Commit, str], parentCommit: Union[C
 
         # commit classification
         if self._commit_classifier and commit.hexsha not in self._clf_results:
-            prob = self._commit_classifier.predict(commit, diff_index)
+            prob = self._commit_classifier.predict(commit, diff_index, self._repo)
             self._clf_results[commit.hexsha] = prob
 
-        # t2: update_graph time
+        # t2: update_graph + git diff traversing time
         t2 = time.monotonic()
+        # t2a: get_contents time
+        t2a = 0
+        # t2a: update_graph time
+        t2b = 0
         if diff_index:
             for diff in diff_index:
                 old_fname, new_fname = _get_fnames(diff)
-                # apply filter
+                # apply file-level filter
                 # if a file comes into/goes from our view, we will set corresponding old_fname/new_fname to None,
                 # as if the file is introduced/removed in this commit.
-                # However, the diff will keep its original, no matter if the file has been filtered in/out.
+                # However, the diff will not change, regardless of whether the file has been filtered out or not.
                 if old_fname and not self._graphServer.filter_file(old_fname):
                     old_fname = None
                 if new_fname and not self._graphServer.filter_file(new_fname):
@@ -254,17 +287,25 @@ async def _analyzeCommit(self, commit: Union[Commit, str], parentCommit: Union[C
 
                 old_src = new_src = None
 
+                t2a0 = time.monotonic()
                 if old_fname:
                     old_src = get_contents(self._repo, parentCommit, old_fname)
+                    if self._file_is_too_large(old_fname, old_src):
+                        continue
 
                 if new_fname:
                     new_src = get_contents(self._repo, commit, new_fname)
+                    if self._file_is_too_large(new_fname, new_src):
+                        continue
+                t2a += time.monotonic() - t2a0
 
+                t2b0 = time.monotonic()
                 if old_src or new_src:
                     result = self._graphServer.update_graph(
                         old_fname, old_src, new_fname, new_src, diff.diff)
                     if asyncio.iscoroutine(result):
                         await result
+                t2b += time.monotonic() - t2b0
         t2 = time.monotonic() - t2
 
         # t3: end_commit time
@@ -275,24 +316,39 @@ async def _analyzeCommit(self, commit: Union[Commit, str], parentCommit: Union[C
         t3 = time.monotonic() - t3
         self._observer.onAfterCommit(self, commit, seekingMode)
         t0 = time.monotonic() - t0
-        _logger.info("t0 = %.2f, t1 = %.2f, t2 = %.2f, t3 = %.2f",
-                     t0, t1, t2, t3)
+        _logger.info("t0 = %.2f, t1 = %.2f, t2 = %.2f, t2a = %.2f, t2b = %.2f, t3 = %.2f",
+                     t0, t1, t2, t2a, t2b, t3)
         assert self._graphServer.get_workspace_commit_hexsha() == commit.hexsha, \
             "GraphServer.get_workspace_commit_hexsha should be return the hexsha seen in last start_commit."
 
     def _filter_monolithic_commit(self, commit: Commit, seeking_mode: CommitSeekingMode) -> CommitSeekingMode:
         # filter monolithic commit
-        if seeking_mode == CommitSeekingMode.NormalForward and len(commit.parents) == 1:
+        # hot fix: enable filter_monolithic_commit on first commit
+        if seeking_mode == CommitSeekingMode.NormalForward and len(commit.parents) <= 1:
             changed_lines = 0
             files = commit.stats.files
             for fname in files:
                 if self._graphServer.filter_file(fname):
                     changed_lines += files[fname]['lines']
+            print('_filter_monolithic_commit commit:', commit.hexsha, 'changed_lines:', changed_lines)
             if changed_lines > self._monolithic_commit_lines_threshold:
                 # enforce using CommitSeekingMode.MergeCommit to update graph without updating node history
+                print('_filter_monolithic_commit set CommitSeekingMode to MergeCommit')
                 return CommitSeekingMode.MergeCommit
         return seeking_mode
 
+    def _file_is_too_large(self, fname, file_content):
+        # Filter monolithic file by its byte size
+        # Returns True if under the threshold
+        file_size = sys.getsizeof(file_content)
+        too_large = file_size > self._monolithic_file_bytes_threshold
+        if too_large:
+            message = 'WARNING: file too large;'
+        else:
+            message = 'OK: file normal size;'
+        print(message, fname, str(file_size / 1000) + 'kB')
+        return too_large
+
 
 def _get_fnames(diff: Diff):
     if diff.new_file:
diff --git a/persper/analytics/call_commit_graph.py b/persper/analytics/call_commit_graph.py
index 0c83fc928c6..48fdf6c2dc2 100644
--- a/persper/analytics/call_commit_graph.py
+++ b/persper/analytics/call_commit_graph.py
@@ -3,13 +3,19 @@
 ====================================
 CallCommitGraph stores all relevant analysis results
 """
+import logging
+import community
 import networkx as nx
 from networkx.readwrite import json_graph
+from typing import Union, Set, List, Dict, Optional
+
 from persper.analytics.devrank import devrank
 from persper.analytics.score import normalize
-from typing import Union, Set, List, Dict, Optional
 from persper.analytics.complexity import eval_project_complexity
 
+_logger = logging.getLogger(__name__)
+
+
 class CommitIdGenerators:
     @staticmethod
     def fromOrdinal(ordinal: int, hexsha: str, message: str):
@@ -50,6 +56,7 @@ def _to_networkx_format(graph_data: Dict) -> Dict:
     def reset(self):
         """Reset all internal states"""
         self._digraph = self._new_graph()
+        self._digraph.degree()
 
     def _new_graph(self):
         """Create a new nx.DiGraph for underlying storage
@@ -97,10 +104,16 @@ def _next_cindex(self):
 
     # TODO: remove the default value of files
     def add_node(self, node: str, files: Union[Set[str], List[str]] = []):
+        if node is None:
+            _logger.error("Argument node is None in add_node.")
+            return
         self._digraph.add_node(node, size=None, history={}, files=set(files))
 
     # add_node must be called on source and target first
     def add_edge(self, source, target):
+        if source is None or target is None:
+            _logger.error("Argument source or target is None in add_edge.")
+            return
         if source not in self._digraph:
             raise ValueError("Error: caller %s does not exist in call-commit graph." % source)
         if target not in self._digraph:
@@ -134,9 +147,15 @@ def update_node_history_accurate(self, node, fstat):
 
     # read/write access to node history are thourgh this function
     def _get_node_history(self, node: str) -> Dict[str, Dict[str, int]]:
+        if node is None:
+            _logger.error("Argument node is None in _get_node_history.")
+            return {}
         return self._digraph.nodes[node]['history']
 
     def update_node_files(self, node: str, new_files: Union[Set[str], List[str]]):
+        if node is None:
+            _logger.error("Argument node is None in update_node_files")
+            return
         self._digraph.nodes[node]['files'] = set(new_files)
 
     # TODO: provide other options for computing a node's size
@@ -148,21 +167,24 @@ def _set_all_nodes_size(self, black_set=None):
         """
         for node in self.nodes():
             node_history = self._get_node_history(node)
-            if black_set is not None:
-                size = 0
-                for cid, chist in node_history.items():
-                    sha = self.commits()[cid]['hexsha']
-                    if sha not in black_set:
-                        size += (chist['adds'] + chist['dels'])
-            else:
-                size = sum([chist['adds'] + chist['dels'] for chist in node_history.values()])
-
+            size = 0
+            for cid, chist in node_history.items():
+                sha = self.commits()[cid]['hexsha']
+                if black_set is not None and sha in black_set:
+                    continue
+                if 'added_units' in chist.keys() and 'removed_units' in chist.keys():
+                    size += (chist['added_units'] + chist['removed_units'])
+                else:
+                    size += (chist['adds'] + chist['dels'])
             # set default size to 1 to avoid zero division error
             if size == 0:
                 size = 1
             self._set_node_size(node, size)
 
     def _set_node_size(self, node, size):
+        if node is None:
+            _logger.error("Argument node is None in _set_node_size.")
+        # set node size even if it is None since we'd like to suppress the error
         self._digraph.nodes[node]['size'] = size
 
     def _set_all_edges_weight(self):
@@ -180,12 +202,17 @@ def eval_project_complexity(self, r_n: float, r_e: float):
         """
         return eval_project_complexity(self._digraph, r_n, r_e)
 
+    def _remove_invalid_nodes(self):
+        if None in self.nodes():
+            self._digraph.remove_node(None)
+
     def function_devranks(self, alpha, black_set=None):
         """
         Args:
                 alpha - A float between 0 and 1, commonly set to 0.85
             black_set - A set of commit hexshas to be blacklisted
         """
+        self._remove_invalid_nodes()
         self._set_all_nodes_size(black_set=black_set)
         return devrank(self._digraph, 'size', alpha=alpha)
 
@@ -206,7 +233,10 @@ def commit_devranks(self, alpha, black_set=None):
                 continue
 
             for cid, chist in history.items():
-                csize = chist['adds'] + chist['dels']
+                if 'added_units' in chist.keys() and 'removed_units' in chist.keys():
+                    csize = (chist['added_units'] + chist['removed_units'])
+                else:
+                    csize = (chist['adds'] + chist['dels'])
                 sha = self.commits()[cid]['hexsha']
                 if black_set is None or sha not in black_set:
                     dr = (csize / size) * func_devranks[func]
@@ -238,3 +268,32 @@ def developer_devranks(self, alpha, black_set=None):
             else:
                 developer_devranks[email] = commit_devranks[sha]
         return developer_devranks
+
+    def compute_modularity(self):
+        """Compute modularity score based on function graph.
+
+        Returns
+        -------
+            modularity : float
+                The modularity score of this graph.
+        """
+        # Check the number of edges
+        if len(self.edges()) == 0:
+            return 0.
+
+        # Construct non directed graph
+        graph = nx.Graph()
+        for node in self.nodes():
+            if node is not None:
+                graph.add_node(node)
+        for (source, target) in self.edges():
+            if source is not None and target is not None:
+                graph.add_edge(source, target)
+        # Compute the partition of the graph nodes
+        partition = community.best_partition(graph)
+        # Compute modularity
+        modularity = community.modularity(partition, graph)
+        # Normalize [0, 1] to [0, 100]
+        modularity = modularity * 100
+
+        return modularity
diff --git a/persper/analytics/call_graph/c.py b/persper/analytics/call_graph/c.py
index 7ff8bbda713..c252e4fc931 100644
--- a/persper/analytics/call_graph/c.py
+++ b/persper/analytics/call_graph/c.py
@@ -18,6 +18,11 @@ class NotFunctionCallError(UnexpectedASTError):
     pass
 
 
+class UnexpectedCallNodeError(UnexpectedASTError):
+    """Raise when failed to parse a function call's callee name"""
+    pass
+
+
 def _handle_function(func_node):
     """Extract name and range from a <function> node
 
@@ -122,17 +127,39 @@ def _handle_call(call_node):
     Case 2: function call from struct variable
         Example: tty->write(tty)
 
+    Case 3: function call in a chain
+        Example: (*mi).second.empty()
+
     Raises:
         NotFunctionCallError
+        UnexpectedCallNodeError
     """
     name_node = call_node.find('srcml:name', ns)
     if name_node is None:
         # Case 1
         raise NotFunctionCallError()
+
+    def last_sub_name_node(node):
+        name_lst = node.findall('srcml:name', ns)
+        if len(name_lst) > 0:
+            return name_lst[-1]
+        else:
+            raise UnexpectedCallNodeError()
+
+    # Case 2 & 3
     callee_name = name_node.text
-    if callee_name is None:
-        # Case 2
-        callee_name = name_node[-1].text
+    # DEBUG
+    # print_flag = False
+    # if callee_name is None:
+    #     print_flag = True
+    #     from persper.analytics.call_graph.utils import transform_node_to_src
+    #     print(transform_node_to_src(name_node))
+    while callee_name is None:
+        name_node = last_sub_name_node(name_node)
+        callee_name = name_node.text
+    # DEBUG
+    # if print_flag:
+    #     print(callee_name)
     return callee_name
 
 
@@ -168,6 +195,9 @@ def update_graph(ccgraph, ast_list, change_stats, new_fname_to_old_fname):
                 except NotFunctionCallError as e:
                     # do not print error since we expect this to happen a lot
                     continue
+                except UnexpectedCallNodeError as e:
+                    print(type(e).__name__, e.args)
+                    continue
 
                 if callee_name not in ccgraph:
                     # Pass [] to files argument since we don't know
diff --git a/persper/analytics/commit_classifier.py b/persper/analytics/commit_classifier.py
index 7e3a5cb58f7..b4bac5d4d1e 100644
--- a/persper/analytics/commit_classifier.py
+++ b/persper/analytics/commit_classifier.py
@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from git import Commit, DiffIndex
+from git import Commit, DiffIndex, Repo
 
 
 class CommitClassifier(ABC):
@@ -8,13 +8,14 @@ class CommitClassifier(ABC):
     """
 
     @abstractmethod
-    def predict(self, commit: Commit, diff_index: DiffIndex):
+    def predict(self, commit: Commit, diff_index: DiffIndex, repo: Repo):
         """
         Args:
             commit: A gitpython's Commit object.
             diff_index: A gitpython's DiffIndex object.
                 It is a list of Diff object, each containing the
                 diff information between a pair of old/new source files.
+            repo: A gitpython's Repo object.
 
 
         Returns:
diff --git a/persper/analytics/git_tools.py b/persper/analytics/git_tools.py
index 01f25e7a8ef..2080a3d0589 100644
--- a/persper/analytics/git_tools.py
+++ b/persper/analytics/git_tools.py
@@ -16,7 +16,10 @@ def diff_with_first_parent(repo: Repo, commit: Commit):
 
 
 def diff_with_commit(repo: Repo, current_commit: Commit, base_commit_sha: str):
-    if not base_commit_sha:
+    # about git.NULL_TREE: https://github.com/gitpython-developers/GitPython/blob/master/git/diff.py#L87
+    if current_commit is None:
+        current_commit = git.NULL_TREE
+    if base_commit_sha is None:
         base_commit = repo.tree(EMPTY_TREE_SHA)
     else:
         base_commit = repo.commit(base_commit_sha)
diff --git a/persper/analytics/graph_server.py b/persper/analytics/graph_server.py
index 406c3f39942..049c721166b 100644
--- a/persper/analytics/graph_server.py
+++ b/persper/analytics/graph_server.py
@@ -6,13 +6,15 @@
 from persper.analytics.call_commit_graph import CallCommitGraph
 
 JS_FILENAME_REGEXES = [
-    r'.+\.js$',
+    r'.+\.(js|vue|ts|tsx)$',
     r'^(?!dist/).+',
     r'^(?!test(s)?/).+',
     r'^(?!spec/).+',
     r'^(?!build/).+',
     r'^(?!bin/).+',
-    r'^(?!doc(s)?/).+'
+    r'^(?!doc(s)?/).+',
+    r'.*(?<!.min.js)$',
+    r'^(?!(.+/)*example(s)?/).+',
 ]
 
 # todo(hezheng) consider moving these regexes to their corresponding language file
@@ -27,7 +29,12 @@
 
 GO_FILENAME_REGEXES = [
     r'.+\.go$',
-    r'^(?!(.+/)*vendor/).+'
+    r'^(?!(.+/)*vendor/).+',
+    r'^(?!Godeps/).+',
+    r'^(?!(.+/)*example(s)?/).+',
+    r'^(?!(.+/)*thrift/gen-go/).+',
+    r'^(?!(.+/)*thrift/thrift-gen/).+',
+    r'^(?!\.gen/go/).+',
 ]
 
 JAVA_FILENAME_REGEXES = [
diff --git a/persper/analytics/multi_analyzer.py b/persper/analytics/multi_analyzer.py
index b6890173c6d..cc1328a1f57 100644
--- a/persper/analytics/multi_analyzer.py
+++ b/persper/analytics/multi_analyzer.py
@@ -1,7 +1,9 @@
 import os
 import json
 import pickle
+import logging
 from git import Repo
+from enum import Enum
 from Naked.toolshed.shell import muterun_rb
 from persper.util.path import root_path
 from persper.analytics.c import CGraphServer
@@ -12,6 +14,23 @@
 from persper.analytics.analyzer2 import Analyzer, AnalyzerObserver, emptyAnalyzerObserver
 
 
+_logger = logging.getLogger(__name__)
+
+
+class Linguist(Enum):
+    """
+    Encodes the relationship between a language and its name in linguist's output
+    https://github.com/github/linguist/blob/master/lib/linguist/languages.yml
+    """
+    C = "C"
+    CPP = "C++"
+    GO = "Go"
+    JAVA = "Java"
+    JAVASCRIPT = "JavaScript"
+    TYPESCRIPT = "TypeScript"
+    VUE = "Vue"
+
+
 class MultiAnalyzer:
 
     LANGUAGE_THRESHOLD = 0.3
@@ -35,14 +54,36 @@ def _set_linguist(self):
             for k in lang_dict.keys():
                 lang_dict[k] = lang_dict[k] * 1.0 / total_lines
 
+            js = Linguist.JAVASCRIPT.value
+            ts = Linguist.TYPESCRIPT.value
+            vue = Linguist.VUE.value
+
+            _logger.info(lang_dict)
+            if vue in lang_dict:
+                _logger.info("Merging Vue to Javascript...")
+                if js not in lang_dict:
+                    lang_dict[js] = 0
+                lang_dict[js] += lang_dict[vue]
+                del lang_dict[vue]
+                _logger.info(lang_dict)
+
+            if ts in lang_dict:
+                _logger.info("Merging TypeScript to JavaScript...")
+                if js not in lang_dict:
+                    lang_dict[js] = 0
+                lang_dict[js] += lang_dict[ts]
+                del lang_dict[ts]
+                _logger.info(lang_dict)
+
+            # intersect with what languages we support
             for lang, value in lang_dict.items():
                 if lang in self._supported_analyzers().keys():
                     self._linguist[lang] = value
 
-            print(self._linguist)
+            _logger.info(self._linguist)
 
         else:
-            print('Analyzing Language Error from Linguist')
+            _logger.info('Analyzing Language Error from Linguist')
 
     def _set_analyzers(self):
         for language, value in self._linguist.items():
@@ -52,7 +93,7 @@ def _set_analyzers(self):
         print(self._analyzers)
 
     async def analyzing(self, saved_path=None):
-        for language, analyzer in self._analyzers.items():    
+        for language, analyzer in self._analyzers.items(): 
             print('Start Analyzing Language, ', language)
             analyzer.terminalCommit = analyzer._repo.head.commit
             await analyzer.analyze()
diff --git a/setup.py b/setup.py
index 34a54934f2d..a079f7e2b88 100644
--- a/setup.py
+++ b/setup.py
@@ -5,5 +5,5 @@
     version = "0.1b1",
     install_requires = ['networkx', 'gitpython', 'numpy', 'scipy', 'requests', 'lxml', 'sklearn', 'nltk', 'openpyxl',
     'pytest', 'pydot', 'ipykernel', 'matplotlib', 'antlr4-python3-runtime', 'python-jsonrpc-server', 'pytest-asyncio',
-    'aenum', 'pytest-cov', 'sphinx' ]
+    'aenum', 'pytest-cov', 'sphinx', 'python-louvain' ]
     )
\ No newline at end of file
diff --git a/test/test_analytics/test_call_commit_graph.py b/test/test_analytics/test_call_commit_graph.py
index aeae9cca858..0097178624f 100644
--- a/test/test_analytics/test_call_commit_graph.py
+++ b/test/test_analytics/test_call_commit_graph.py
@@ -95,6 +95,33 @@ def test_call_commit_graph():
     assert isclose(dev_drs3[second_commit['authorEmail']], 0.201, rel_tol=1e-2)
 
 
+def test_devrank_with_accurate_history():
+    ccgraph = CallCommitGraph()
+    first_commit = {
+        'hexsha': '0x01',
+        'authorName': 'koala',
+        'authorEmail': 'koala@persper.org',
+        'message': 'first commit'
+    }
+    ccgraph.add_commit(first_commit['hexsha'],
+                       first_commit['authorName'],
+                       first_commit['authorEmail'],
+                       first_commit['message'])
+    ccgraph.add_node('f1')
+    ccgraph.update_node_history_accurate('f1', {'adds': 10, 'dels': 0, 'added_units': 20, 'removed_units': 0})
+    ccgraph.add_node('f2')
+    ccgraph.update_node_history_accurate('f2', {'adds': 10, 'dels': 0, 'added_units': 40, 'removed_units': 0})
+    ccgraph.add_edge('f1', 'f2')
+
+    func_drs = ccgraph.function_devranks(0.85)
+    commit_drs = ccgraph.commit_devranks(0.85)
+    dev_drs = ccgraph.developer_devranks(0.85)
+    assert isclose(func_drs['f1'], 0.26, rel_tol=1e-2)
+    assert isclose(func_drs['f2'], 0.74, rel_tol=1e-2)
+    assert isclose(commit_drs[first_commit['hexsha']], 1)
+    assert isclose(dev_drs[first_commit['authorEmail']], 1)
+
+
 @pytest.mark.asyncio
 async def test_black_set():
     """
@@ -121,3 +148,13 @@ async def test_black_set():
     assert len(devdict2) == 2
     assert parent_sha in devdict
     assert parent_sha not in devdict2
+
+
+def test_remove_invalid_nodes():
+    ccgraph = CallCommitGraph()
+    ccgraph.add_node('f1')
+    ccgraph.add_node(None)
+    ccgraph.add_edge('f1', None)
+
+    func_drs = ccgraph.function_devranks(0.85)
+    assert isclose(func_drs['f1'], 1, rel_tol=1e-2)
diff --git a/test/test_analytics/test_diff.py b/test/test_analytics/test_diff.py
index fc07cbb9708..0aeeb5ba6a4 100644
--- a/test/test_analytics/test_diff.py
+++ b/test/test_analytics/test_diff.py
@@ -1,22 +1,40 @@
 import os
+import pytest
 from git import Repo
-from persper.analytics.git_tools import diff_with_first_parent
+from persper.analytics.git_tools import diff_with_first_parent, diff_with_commit
 from persper.util.path import root_path
 
 
-def test_diff_ignore_space():
+@pytest.fixture(scope='module')
+def bitcoin_repo():
+    repo_path = os.path.join(root_path, "repos/bitcoin")
+    bitcoin_url = 'https://github.com/bitcoin/bitcoin'
+    if not os.path.exists(repo_path):
+        Repo.clone_from(bitcoin_url, repo_path)
+    return Repo(repo_path)
+
+
+def test_diff_ignore_space(bitcoin_repo):
     """
     bitcoin project has a commit which only converts CRLF to LF
     its diff with parent should be empty when
     ignore space option is enabled
     The CRLF commit: https://github.com/bitcoin/bitcoin/commit/0a61b0df1224a5470bcddab302bc199ca5a9e356
     """
-    repo_path = os.path.join(root_path, "repos/bitcoin")
-    bitcoin_url = 'https://github.com/bitcoin/bitcoin'
-    if not os.path.exists(repo_path):
-        Repo.clone_from(bitcoin_url, repo_path)
-    r = Repo(repo_path)
     crlf_sha = '0a61b0df1224a5470bcddab302bc199ca5a9e356'
-    crlf_commit = r.commit(crlf_sha)
-    diff_result = diff_with_first_parent(r, crlf_commit)
+    crlf_commit = bitcoin_repo.commit(crlf_sha)
+    diff_result = diff_with_first_parent(bitcoin_repo, crlf_commit)
     assert len(diff_result) == 0
+
+
+def test_empty_current_commit(bitcoin_repo):
+    """
+    When rewinding to an orphaned commit, the `current_commit` passed to the `diff_with_commit` function is None.
+    This test case makes sure we handle this scenario without throwing an `Exception`.
+    `base_commit` is the first commit in the bitcoin repo, which adds 45 new files
+    Link: https://github.com/bitcoin/bitcoin/commit/4405b78d6059e536c36974088a8ed4d9f0f29898
+    """
+    current_commit = None
+    base_commit_sha = '4405b78d6059e536c36974088a8ed4d9f0f29898'
+    diff_index = diff_with_commit(bitcoin_repo, current_commit, base_commit_sha)
+    assert len(diff_index) == 45
diff --git a/test/test_analytics/test_filter_commit.py b/test/test_analytics/test_filter_commit.py
index 802ffbaddb9..785c41627d7 100644
--- a/test/test_analytics/test_filter_commit.py
+++ b/test/test_analytics/test_filter_commit.py
@@ -29,13 +29,13 @@ def test_analyzer_filter_monolithic_commit(az):
     threshold = az._monolithic_commit_lines_threshold
 
     # case 1: changes above threshold, but the commit is the first commit
-    # expected result: normal forward
+    # expected result: merge commit
     case_1_files = {
         'main.c': {'lines': threshold + 1},
     }
     case_1_commit = MockCommit(case_1_files, 0)
     case_1_seeking_mode = az._filter_monolithic_commit(case_1_commit, CommitSeekingMode.NormalForward)
-    assert case_1_seeking_mode == CommitSeekingMode.NormalForward
+    assert case_1_seeking_mode == CommitSeekingMode.MergeCommit
 
     # case 2: changes equal to threshold, the commit has one parent commit
     # expected result: normal forward
@@ -68,6 +68,7 @@ def test_analyzer_filter_monolithic_commit(az):
 
 class MockCommit:
     def __init__(self, files: dict, parent_number: int = 1):
+        self.hexsha = 'test'
         self.stats = MockCommitStats(files)
         self.parents = [{}] * parent_number
 
diff --git a/test/test_analytics/test_modularity.py b/test/test_analytics/test_modularity.py
new file mode 100644
index 00000000000..db0d314eda8
--- /dev/null
+++ b/test/test_analytics/test_modularity.py
@@ -0,0 +1,20 @@
+# -*- coding: utf-8 -*-
+
+from persper.analytics.call_commit_graph import CallCommitGraph
+
+
+def test_modularity():
+    ccgraph = CallCommitGraph()
+    for i in range(0, 100):
+        ccgraph.add_node("node" + str(i))
+    # No any edges
+    assert ccgraph.compute_modularity() == 0
+
+    # Only one edge
+    ccgraph.add_edge("node0", "node1")
+    assert ccgraph.compute_modularity() == 0
+
+    # Multiple edges
+    for i in range(1, 99):
+        ccgraph.add_edge("node" + str(i), "node" + str(i + 1))
+    assert int(ccgraph.compute_modularity()) == 80
diff --git a/test/test_analytics/test_multi_analyzer.py b/test/test_analytics/test_multi_analyzer.py
new file mode 100644
index 00000000000..c2bbaa6ae56
--- /dev/null
+++ b/test/test_analytics/test_multi_analyzer.py
@@ -0,0 +1,33 @@
+import os
+import pytest
+from git import Repo
+from persper.analytics.multi_analyzer import MultiAnalyzer
+from persper.util.path import root_path
+
+
+@pytest.fixture(scope='module')
+def vue_repo_path():
+    repo_path = os.path.join(root_path, "repos/vue-realworld-example-app")
+    repo_url = 'https://github.com/gothinkster/vue-realworld-example-app'
+    if not os.path.exists(repo_path):
+        Repo.clone_from(repo_url, repo_path)
+    return repo_path
+
+
+@pytest.fixture(scope='module')
+def ts_repo_path():
+    repo_path = os.path.join(root_path, "repos/TypeScriptSamples")
+    repo_url = 'https://github.com/microsoft/TypeScriptSamples'
+    if not os.path.exists(repo_path):
+        Repo.clone_from(repo_url, repo_path)
+    return repo_path
+
+
+def test_set_linguist_vue(vue_repo_path):
+    # _set_linguist is called during the initialization of MultiAnalyzer
+    maz = MultiAnalyzer(vue_repo_path)
+
+
+def test_set_linguist_ts(ts_repo_path):
+    # _set_linguist is called during the initialization of MultiAnalyzer
+    maz = MultiAnalyzer(ts_repo_path)