diff --git a/tools/rw-heatmaps/README.md b/tools/rw-heatmaps/README.md
index 893ea9871cd..2d4b0fada69 100644
--- a/tools/rw-heatmaps/README.md
+++ b/tools/rw-heatmaps/README.md
@@ -15,6 +15,12 @@ To get a mixed read/write performance evaluation result:
 Note: the result csv file will be saved to current working directory. The working directory is where etcd database is saved. The working directory is designed for scenarios where a different mounted disk is preferred.
 
 ### Plot Graphs
+Get the dependencies via pip:
+
+```sh
+pip3 install -r requirements.txt
+```
+
 To generate two images (read and write) based on the benchmark result csv file:
 ```sh
 # to generate a pair of read & write images from one data csv file
diff --git a/tools/rw-heatmaps/plot_data.py b/tools/rw-heatmaps/plot_data.py
index 217eb6f8a8b..e4116663d63 100755
--- a/tools/rw-heatmaps/plot_data.py
+++ b/tools/rw-heatmaps/plot_data.py
@@ -78,107 +78,6 @@ def load_data_files(*args):
         sys.exit(1)
     return res
 
-
-# This is copied directly from matplotlib source code. Some early versions of matplotlib
-# do not have CenteredNorm class
-class CenteredNorm(colors.Normalize):
-
-    def __init__(self, vcenter=0, halfrange=None, clip=False):
-        """
-        Normalize symmetrical data around a center (0 by default).
-
-        Unlike `TwoSlopeNorm`, `CenteredNorm` applies an equal rate of change
-        around the center.
-
-        Useful when mapping symmetrical data around a conceptual center
-        e.g., data that range from -2 to 4, with 0 as the midpoint, and
-        with equal rates of change around that midpoint.
-
-        Parameters
-        ----------
-        vcenter : float, default: 0
-            The data value that defines ``0.5`` in the normalization.
-        halfrange : float, optional
-            The range of data values that defines a range of ``0.5`` in the
-            normalization, so that *vcenter* - *halfrange* is ``0.0`` and
-            *vcenter* + *halfrange* is ``1.0`` in the normalization.
-            Defaults to the largest absolute difference to *vcenter* for
-            the values in the dataset.
-
-        Examples
-        --------
-        This maps data values -2 to 0.25, 0 to 0.5, and 4 to 1.0
-        (assuming equal rates of change above and below 0.0):
-
-            >>> import matplotlib.colors as mcolors
-            >>> norm = mcolors.CenteredNorm(halfrange=4.0)
-            >>> data = [-2., 0., 4.]
-            >>> norm(data)
-            array([0.25, 0.5 , 1.  ])
-        """
-        self._vcenter = vcenter
-        self.vmin = None
-        self.vmax = None
-        # calling the halfrange setter to set vmin and vmax
-        self.halfrange = halfrange
-        self.clip = clip
-
-    def _set_vmin_vmax(self):
-        """
-        Set *vmin* and *vmax* based on *vcenter* and *halfrange*.
-        """
-        self.vmax = self._vcenter + self._halfrange
-        self.vmin = self._vcenter - self._halfrange
-
-    def autoscale(self, A):
-        """
-        Set *halfrange* to ``max(abs(A-vcenter))``, then set *vmin* and *vmax*.
-        """
-        A = np.asanyarray(A)
-        self._halfrange = max(self._vcenter-A.min(),
-                              A.max()-self._vcenter)
-        self._set_vmin_vmax()
-
-    def autoscale_None(self, A):
-        """Set *vmin* and *vmax*."""
-        A = np.asanyarray(A)
-        if self._halfrange is None and A.size:
-            self.autoscale(A)
-
-    @property
-    def vcenter(self):
-        return self._vcenter
-
-    @vcenter.setter
-    def vcenter(self, vcenter):
-        self._vcenter = vcenter
-        if self.vmax is not None:
-            # recompute halfrange assuming vmin and vmax represent
-            # min and max of data
-            self._halfrange = max(self._vcenter-self.vmin,
-                                  self.vmax-self._vcenter)
-            self._set_vmin_vmax()
-
-    @property
-    def halfrange(self):
-        return self._halfrange
-
-    @halfrange.setter
-    def halfrange(self, halfrange):
-        if halfrange is None:
-            self._halfrange = None
-            self.vmin = None
-            self.vmax = None
-        else:
-            self._halfrange = abs(halfrange)
-
-    def __call__(self, value, clip=None):
-        if self._halfrange is not None:
-            # enforce symmetry, reset vmin and vmax
-            self._set_vmin_vmax()
-        return super().__call__(value, clip=clip)
-
-
 # plot type is the type of the data to plot. Either 'read' or 'write'
 def plot_data(title, plot_type, cmap_name_default, *args):
     if len(args) == 1:
@@ -220,7 +119,7 @@ def plot_data(title, plot_type, cmap_name_default, *args):
                 if col == 2:
                     cmap_name = 'bwr'
                     if params.zero:
-                        norm = CenteredNorm()
+                        norm = colors.CenteredNorm()
                 else:
                     cmap_name = cmap_name_default
                 plt.tripcolor(df['conn_size'], df['value_size'], df[plot_type],
diff --git a/tools/rw-heatmaps/requirements.txt b/tools/rw-heatmaps/requirements.txt
new file mode 100644
index 00000000000..c8048003bfc
--- /dev/null
+++ b/tools/rw-heatmaps/requirements.txt
@@ -0,0 +1,3 @@
+matplotlib>=3.6
+numpy>=1.23
+pandas>=1.5